├── .gitattributes ├── .github └── ISSUE_TEMPLATE │ └── bug_report.md ├── .gitignore ├── CONTRIBUTING.md ├── LICENSE ├── README.md ├── locales ├── de.json ├── es.json ├── fi.json ├── km.json ├── ko.json ├── pt-br.json ├── ru.json ├── th.json └── zh.json ├── media ├── Download.png ├── avgpcperformancedemo.gif ├── deepwarebench.gif ├── demo.gif ├── instruction.png ├── live_show.gif ├── ludwig.gif ├── meme.gif ├── movie.gif └── streamers.gif ├── models └── instructions.txt ├── modules ├── __init__.py ├── capturer.py ├── cluster_analysis.py ├── core.py ├── face_analyser.py ├── gettext.py ├── globals.py ├── metadata.py ├── predicter.py ├── processors │ ├── __init__.py │ └── frame │ │ ├── __init__.py │ │ ├── core.py │ │ ├── face_enhancer.py │ │ └── face_swapper.py ├── typing.py ├── ui.json ├── ui.py ├── utilities.py └── video_capture.py ├── mypi.ini ├── requirements.txt ├── run-cuda.bat ├── run-directml.bat └── run.py /.gitattributes: -------------------------------------------------------------------------------- 1 | # Auto detect text files and perform LF normalization 2 | * text=auto 3 | -------------------------------------------------------------------------------- /.github/ISSUE_TEMPLATE/bug_report.md: -------------------------------------------------------------------------------- 1 | ***[Remove this]The issue would be closed without notice and be considered spam if the template is not followed.*** 2 | 3 | **Describe the bug** 4 | A clear and concise description of what the bug is. 5 | 6 | **Screenshots** 7 | If applicable, add screenshots to help explain your problem. 8 | 9 | **Error Message** 10 | 11 | `` 12 | 13 | **Desktop (please complete the following information):** 14 | - OS: [e.g. Windows] 15 | - Version [e.g. 22] 16 | - GPU 17 | - CPU 18 | 19 | **Additional context** 20 | Add any other context about the problem here. 21 | 22 | **Confirmation (Mandatory)** 23 | - [ ] I have followed the template 24 | - [ ] This is not a query about how to increase performance 25 | - [ ] I have checked the issues page, and this is not a duplicate 26 | 27 | -------------------------------------------------------------------------------- /.gitignore: -------------------------------------------------------------------------------- 1 | __pycache__/ 2 | *.py[cod] 3 | *$py.class 4 | *.pyc 5 | .idea 6 | .todo 7 | *.log 8 | *.backup 9 | tf_env/ 10 | *.png 11 | *.mp4 12 | *.mkv 13 | 14 | .tmp/ 15 | temp/ 16 | .venv/ 17 | venv/ 18 | env/ 19 | workflow/ 20 | gfpgan/ 21 | models/inswapper_128.onnx 22 | models/GFPGANv1.4.pth 23 | *.onnx 24 | models/DMDNet.pth 25 | faceswap/ 26 | .vscode/ 27 | switch_states.json 28 | -------------------------------------------------------------------------------- /CONTRIBUTING.md: -------------------------------------------------------------------------------- 1 | # Collaboration Guidelines and Codebase Quality Standards 2 | 3 | To ensure smooth collaboration and maintain the high quality of our codebase, please adhere to the following guidelines: 4 | 5 | ## Branching Strategy 6 | 7 | * **`premain`**: 8 | * Always push your changes to the `premain` branch initially. 9 | * This safeguards the `main` branch from unintentional disruptions. 10 | * All tests will be performed on the `premain` branch. 11 | * Changes will only be merged into `main` after several hours or days of rigorous testing. 12 | * **`experimental`**: 13 | * For large or potentially disruptive changes, use the `experimental` branch. 14 | * This allows for thorough discussion and review before considering a merge into `main`. 15 | 16 | ## Pre-Pull Request Checklist 17 | 18 | Before creating a Pull Request (PR), ensure you have completed the following tests: 19 | 20 | ### Functionality 21 | 22 | * **Realtime Faceswap**: 23 | * Test with face enhancer **enabled** and **disabled**. 24 | * **Map Faces**: 25 | * Test with both options (**enabled** and **disabled**). 26 | * **Camera Listing**: 27 | * Verify that all cameras are listed accurately. 28 | 29 | ### Stability 30 | 31 | * **Realtime FPS**: 32 | * Confirm that there is no drop in real-time frames per second (FPS). 33 | * **Boot Time**: 34 | * Changes should not negatively impact the boot time of either the application or the real-time faceswap feature. 35 | * **GPU Overloading**: 36 | * Test for a minimum of 15 minutes to guarantee no GPU overloading, which could lead to crashes. 37 | * **App Performance**: 38 | * The application should remain responsive and not exhibit any lag. 39 | -------------------------------------------------------------------------------- /LICENSE: -------------------------------------------------------------------------------- 1 | GNU AFFERO GENERAL PUBLIC LICENSE 2 | Version 3, 19 November 2007 3 | 4 | Copyright (C) 2007 Free Software Foundation, Inc. 5 | Everyone is permitted to copy and distribute verbatim copies 6 | of this license document, but changing it is not allowed. 7 | 8 | Preamble 9 | 10 | The GNU Affero General Public License is a free, copyleft license for 11 | software and other kinds of works, specifically designed to ensure 12 | cooperation with the community in the case of network server software. 13 | 14 | The licenses for most software and other practical works are designed 15 | to take away your freedom to share and change the works. By contrast, 16 | our General Public Licenses are intended to guarantee your freedom to 17 | share and change all versions of a program--to make sure it remains free 18 | software for all its users. 19 | 20 | When we speak of free software, we are referring to freedom, not 21 | price. Our General Public Licenses are designed to make sure that you 22 | have the freedom to distribute copies of free software (and charge for 23 | them if you wish), that you receive source code or can get it if you 24 | want it, that you can change the software or use pieces of it in new 25 | free programs, and that you know you can do these things. 26 | 27 | Developers that use our General Public Licenses protect your rights 28 | with two steps: (1) assert copyright on the software, and (2) offer 29 | you this License which gives you legal permission to copy, distribute 30 | and/or modify the software. 31 | 32 | A secondary benefit of defending all users' freedom is that 33 | improvements made in alternate versions of the program, if they 34 | receive widespread use, become available for other developers to 35 | incorporate. Many developers of free software are heartened and 36 | encouraged by the resulting cooperation. However, in the case of 37 | software used on network servers, this result may fail to come about. 38 | The GNU General Public License permits making a modified version and 39 | letting the public access it on a server without ever releasing its 40 | source code to the public. 41 | 42 | The GNU Affero General Public License is designed specifically to 43 | ensure that, in such cases, the modified source code becomes available 44 | to the community. It requires the operator of a network server to 45 | provide the source code of the modified version running there to the 46 | users of that server. Therefore, public use of a modified version, on 47 | a publicly accessible server, gives the public access to the source 48 | code of the modified version. 49 | 50 | An older license, called the Affero General Public License and 51 | published by Affero, was designed to accomplish similar goals. This is 52 | a different license, not a version of the Affero GPL, but Affero has 53 | released a new version of the Affero GPL which permits relicensing under 54 | this license. 55 | 56 | The precise terms and conditions for copying, distribution and 57 | modification follow. 58 | 59 | TERMS AND CONDITIONS 60 | 61 | 0. Definitions. 62 | 63 | "This License" refers to version 3 of the GNU Affero General Public License. 64 | 65 | "Copyright" also means copyright-like laws that apply to other kinds of 66 | works, such as semiconductor masks. 67 | 68 | "The Program" refers to any copyrightable work licensed under this 69 | License. Each licensee is addressed as "you". "Licensees" and 70 | "recipients" may be individuals or organizations. 71 | 72 | To "modify" a work means to copy from or adapt all or part of the work 73 | in a fashion requiring copyright permission, other than the making of an 74 | exact copy. The resulting work is called a "modified version" of the 75 | earlier work or a work "based on" the earlier work. 76 | 77 | A "covered work" means either the unmodified Program or a work based 78 | on the Program. 79 | 80 | To "propagate" a work means to do anything with it that, without 81 | permission, would make you directly or secondarily liable for 82 | infringement under applicable copyright law, except executing it on a 83 | computer or modifying a private copy. Propagation includes copying, 84 | distribution (with or without modification), making available to the 85 | public, and in some countries other activities as well. 86 | 87 | To "convey" a work means any kind of propagation that enables other 88 | parties to make or receive copies. Mere interaction with a user through 89 | a computer network, with no transfer of a copy, is not conveying. 90 | 91 | An interactive user interface displays "Appropriate Legal Notices" 92 | to the extent that it includes a convenient and prominently visible 93 | feature that (1) displays an appropriate copyright notice, and (2) 94 | tells the user that there is no warranty for the work (except to the 95 | extent that warranties are provided), that licensees may convey the 96 | work under this License, and how to view a copy of this License. If 97 | the interface presents a list of user commands or options, such as a 98 | menu, a prominent item in the list meets this criterion. 99 | 100 | 1. Source Code. 101 | 102 | The "source code" for a work means the preferred form of the work 103 | for making modifications to it. "Object code" means any non-source 104 | form of a work. 105 | 106 | A "Standard Interface" means an interface that either is an official 107 | standard defined by a recognized standards body, or, in the case of 108 | interfaces specified for a particular programming language, one that 109 | is widely used among developers working in that language. 110 | 111 | The "System Libraries" of an executable work include anything, other 112 | than the work as a whole, that (a) is included in the normal form of 113 | packaging a Major Component, but which is not part of that Major 114 | Component, and (b) serves only to enable use of the work with that 115 | Major Component, or to implement a Standard Interface for which an 116 | implementation is available to the public in source code form. A 117 | "Major Component", in this context, means a major essential component 118 | (kernel, window system, and so on) of the specific operating system 119 | (if any) on which the executable work runs, or a compiler used to 120 | produce the work, or an object code interpreter used to run it. 121 | 122 | The "Corresponding Source" for a work in object code form means all 123 | the source code needed to generate, install, and (for an executable 124 | work) run the object code and to modify the work, including scripts to 125 | control those activities. However, it does not include the work's 126 | System Libraries, or general-purpose tools or generally available free 127 | programs which are used unmodified in performing those activities but 128 | which are not part of the work. For example, Corresponding Source 129 | includes interface definition files associated with source files for 130 | the work, and the source code for shared libraries and dynamically 131 | linked subprograms that the work is specifically designed to require, 132 | such as by intimate data communication or control flow between those 133 | subprograms and other parts of the work. 134 | 135 | The Corresponding Source need not include anything that users 136 | can regenerate automatically from other parts of the Corresponding 137 | Source. 138 | 139 | The Corresponding Source for a work in source code form is that 140 | same work. 141 | 142 | 2. Basic Permissions. 143 | 144 | All rights granted under this License are granted for the term of 145 | copyright on the Program, and are irrevocable provided the stated 146 | conditions are met. This License explicitly affirms your unlimited 147 | permission to run the unmodified Program. The output from running a 148 | covered work is covered by this License only if the output, given its 149 | content, constitutes a covered work. This License acknowledges your 150 | rights of fair use or other equivalent, as provided by copyright law. 151 | 152 | You may make, run and propagate covered works that you do not 153 | convey, without conditions so long as your license otherwise remains 154 | in force. You may convey covered works to others for the sole purpose 155 | of having them make modifications exclusively for you, or provide you 156 | with facilities for running those works, provided that you comply with 157 | the terms of this License in conveying all material for which you do 158 | not control copyright. Those thus making or running the covered works 159 | for you must do so exclusively on your behalf, under your direction 160 | and control, on terms that prohibit them from making any copies of 161 | your copyrighted material outside their relationship with you. 162 | 163 | Conveying under any other circumstances is permitted solely under 164 | the conditions stated below. Sublicensing is not allowed; section 10 165 | makes it unnecessary. 166 | 167 | 3. Protecting Users' Legal Rights From Anti-Circumvention Law. 168 | 169 | No covered work shall be deemed part of an effective technological 170 | measure under any applicable law fulfilling obligations under article 171 | 11 of the WIPO copyright treaty adopted on 20 December 1996, or 172 | similar laws prohibiting or restricting circumvention of such 173 | measures. 174 | 175 | When you convey a covered work, you waive any legal power to forbid 176 | circumvention of technological measures to the extent such circumvention 177 | is effected by exercising rights under this License with respect to 178 | the covered work, and you disclaim any intention to limit operation or 179 | modification of the work as a means of enforcing, against the work's 180 | users, your or third parties' legal rights to forbid circumvention of 181 | technological measures. 182 | 183 | 4. Conveying Verbatim Copies. 184 | 185 | You may convey verbatim copies of the Program's source code as you 186 | receive it, in any medium, provided that you conspicuously and 187 | appropriately publish on each copy an appropriate copyright notice; 188 | keep intact all notices stating that this License and any 189 | non-permissive terms added in accord with section 7 apply to the code; 190 | keep intact all notices of the absence of any warranty; and give all 191 | recipients a copy of this License along with the Program. 192 | 193 | You may charge any price or no price for each copy that you convey, 194 | and you may offer support or warranty protection for a fee. 195 | 196 | 5. Conveying Modified Source Versions. 197 | 198 | You may convey a work based on the Program, or the modifications to 199 | produce it from the Program, in the form of source code under the 200 | terms of section 4, provided that you also meet all of these conditions: 201 | 202 | a) The work must carry prominent notices stating that you modified 203 | it, and giving a relevant date. 204 | 205 | b) The work must carry prominent notices stating that it is 206 | released under this License and any conditions added under section 207 | 7. This requirement modifies the requirement in section 4 to 208 | "keep intact all notices". 209 | 210 | c) You must license the entire work, as a whole, under this 211 | License to anyone who comes into possession of a copy. This 212 | License will therefore apply, along with any applicable section 7 213 | additional terms, to the whole of the work, and all its parts, 214 | regardless of how they are packaged. This License gives no 215 | permission to license the work in any other way, but it does not 216 | invalidate such permission if you have separately received it. 217 | 218 | d) If the work has interactive user interfaces, each must display 219 | Appropriate Legal Notices; however, if the Program has interactive 220 | interfaces that do not display Appropriate Legal Notices, your 221 | work need not make them do so. 222 | 223 | A compilation of a covered work with other separate and independent 224 | works, which are not by their nature extensions of the covered work, 225 | and which are not combined with it such as to form a larger program, 226 | in or on a volume of a storage or distribution medium, is called an 227 | "aggregate" if the compilation and its resulting copyright are not 228 | used to limit the access or legal rights of the compilation's users 229 | beyond what the individual works permit. Inclusion of a covered work 230 | in an aggregate does not cause this License to apply to the other 231 | parts of the aggregate. 232 | 233 | 6. Conveying Non-Source Forms. 234 | 235 | You may convey a covered work in object code form under the terms 236 | of sections 4 and 5, provided that you also convey the 237 | machine-readable Corresponding Source under the terms of this License, 238 | in one of these ways: 239 | 240 | a) Convey the object code in, or embodied in, a physical product 241 | (including a physical distribution medium), accompanied by the 242 | Corresponding Source fixed on a durable physical medium 243 | customarily used for software interchange. 244 | 245 | b) Convey the object code in, or embodied in, a physical product 246 | (including a physical distribution medium), accompanied by a 247 | written offer, valid for at least three years and valid for as 248 | long as you offer spare parts or customer support for that product 249 | model, to give anyone who possesses the object code either (1) a 250 | copy of the Corresponding Source for all the software in the 251 | product that is covered by this License, on a durable physical 252 | medium customarily used for software interchange, for a price no 253 | more than your reasonable cost of physically performing this 254 | conveying of source, or (2) access to copy the 255 | Corresponding Source from a network server at no charge. 256 | 257 | c) Convey individual copies of the object code with a copy of the 258 | written offer to provide the Corresponding Source. This 259 | alternative is allowed only occasionally and noncommercially, and 260 | only if you received the object code with such an offer, in accord 261 | with subsection 6b. 262 | 263 | d) Convey the object code by offering access from a designated 264 | place (gratis or for a charge), and offer equivalent access to the 265 | Corresponding Source in the same way through the same place at no 266 | further charge. You need not require recipients to copy the 267 | Corresponding Source along with the object code. If the place to 268 | copy the object code is a network server, the Corresponding Source 269 | may be on a different server (operated by you or a third party) 270 | that supports equivalent copying facilities, provided you maintain 271 | clear directions next to the object code saying where to find the 272 | Corresponding Source. Regardless of what server hosts the 273 | Corresponding Source, you remain obligated to ensure that it is 274 | available for as long as needed to satisfy these requirements. 275 | 276 | e) Convey the object code using peer-to-peer transmission, provided 277 | you inform other peers where the object code and Corresponding 278 | Source of the work are being offered to the general public at no 279 | charge under subsection 6d. 280 | 281 | A separable portion of the object code, whose source code is excluded 282 | from the Corresponding Source as a System Library, need not be 283 | included in conveying the object code work. 284 | 285 | A "User Product" is either (1) a "consumer product", which means any 286 | tangible personal property which is normally used for personal, family, 287 | or household purposes, or (2) anything designed or sold for incorporation 288 | into a dwelling. In determining whether a product is a consumer product, 289 | doubtful cases shall be resolved in favor of coverage. For a particular 290 | product received by a particular user, "normally used" refers to a 291 | typical or common use of that class of product, regardless of the status 292 | of the particular user or of the way in which the particular user 293 | actually uses, or expects or is expected to use, the product. A product 294 | is a consumer product regardless of whether the product has substantial 295 | commercial, industrial or non-consumer uses, unless such uses represent 296 | the only significant mode of use of the product. 297 | 298 | "Installation Information" for a User Product means any methods, 299 | procedures, authorization keys, or other information required to install 300 | and execute modified versions of a covered work in that User Product from 301 | a modified version of its Corresponding Source. The information must 302 | suffice to ensure that the continued functioning of the modified object 303 | code is in no case prevented or interfered with solely because 304 | modification has been made. 305 | 306 | If you convey an object code work under this section in, or with, or 307 | specifically for use in, a User Product, and the conveying occurs as 308 | part of a transaction in which the right of possession and use of the 309 | User Product is transferred to the recipient in perpetuity or for a 310 | fixed term (regardless of how the transaction is characterized), the 311 | Corresponding Source conveyed under this section must be accompanied 312 | by the Installation Information. But this requirement does not apply 313 | if neither you nor any third party retains the ability to install 314 | modified object code on the User Product (for example, the work has 315 | been installed in ROM). 316 | 317 | The requirement to provide Installation Information does not include a 318 | requirement to continue to provide support service, warranty, or updates 319 | for a work that has been modified or installed by the recipient, or for 320 | the User Product in which it has been modified or installed. Access to a 321 | network may be denied when the modification itself materially and 322 | adversely affects the operation of the network or violates the rules and 323 | protocols for communication across the network. 324 | 325 | Corresponding Source conveyed, and Installation Information provided, 326 | in accord with this section must be in a format that is publicly 327 | documented (and with an implementation available to the public in 328 | source code form), and must require no special password or key for 329 | unpacking, reading or copying. 330 | 331 | 7. Additional Terms. 332 | 333 | "Additional permissions" are terms that supplement the terms of this 334 | License by making exceptions from one or more of its conditions. 335 | Additional permissions that are applicable to the entire Program shall 336 | be treated as though they were included in this License, to the extent 337 | that they are valid under applicable law. If additional permissions 338 | apply only to part of the Program, that part may be used separately 339 | under those permissions, but the entire Program remains governed by 340 | this License without regard to the additional permissions. 341 | 342 | When you convey a copy of a covered work, you may at your option 343 | remove any additional permissions from that copy, or from any part of 344 | it. (Additional permissions may be written to require their own 345 | removal in certain cases when you modify the work.) You may place 346 | additional permissions on material, added by you to a covered work, 347 | for which you have or can give appropriate copyright permission. 348 | 349 | Notwithstanding any other provision of this License, for material you 350 | add to a covered work, you may (if authorized by the copyright holders of 351 | that material) supplement the terms of this License with terms: 352 | 353 | a) Disclaiming warranty or limiting liability differently from the 354 | terms of sections 15 and 16 of this License; or 355 | 356 | b) Requiring preservation of specified reasonable legal notices or 357 | author attributions in that material or in the Appropriate Legal 358 | Notices displayed by works containing it; or 359 | 360 | c) Prohibiting misrepresentation of the origin of that material, or 361 | requiring that modified versions of such material be marked in 362 | reasonable ways as different from the original version; or 363 | 364 | d) Limiting the use for publicity purposes of names of licensors or 365 | authors of the material; or 366 | 367 | e) Declining to grant rights under trademark law for use of some 368 | trade names, trademarks, or service marks; or 369 | 370 | f) Requiring indemnification of licensors and authors of that 371 | material by anyone who conveys the material (or modified versions of 372 | it) with contractual assumptions of liability to the recipient, for 373 | any liability that these contractual assumptions directly impose on 374 | those licensors and authors. 375 | 376 | All other non-permissive additional terms are considered "further 377 | restrictions" within the meaning of section 10. If the Program as you 378 | received it, or any part of it, contains a notice stating that it is 379 | governed by this License along with a term that is a further 380 | restriction, you may remove that term. If a license document contains 381 | a further restriction but permits relicensing or conveying under this 382 | License, you may add to a covered work material governed by the terms 383 | of that license document, provided that the further restriction does 384 | not survive such relicensing or conveying. 385 | 386 | If you add terms to a covered work in accord with this section, you 387 | must place, in the relevant source files, a statement of the 388 | additional terms that apply to those files, or a notice indicating 389 | where to find the applicable terms. 390 | 391 | Additional terms, permissive or non-permissive, may be stated in the 392 | form of a separately written license, or stated as exceptions; 393 | the above requirements apply either way. 394 | 395 | 8. Termination. 396 | 397 | You may not propagate or modify a covered work except as expressly 398 | provided under this License. Any attempt otherwise to propagate or 399 | modify it is void, and will automatically terminate your rights under 400 | this License (including any patent licenses granted under the third 401 | paragraph of section 11). 402 | 403 | However, if you cease all violation of this License, then your 404 | license from a particular copyright holder is reinstated (a) 405 | provisionally, unless and until the copyright holder explicitly and 406 | finally terminates your license, and (b) permanently, if the copyright 407 | holder fails to notify you of the violation by some reasonable means 408 | prior to 60 days after the cessation. 409 | 410 | Moreover, your license from a particular copyright holder is 411 | reinstated permanently if the copyright holder notifies you of the 412 | violation by some reasonable means, this is the first time you have 413 | received notice of violation of this License (for any work) from that 414 | copyright holder, and you cure the violation prior to 30 days after 415 | your receipt of the notice. 416 | 417 | Termination of your rights under this section does not terminate the 418 | licenses of parties who have received copies or rights from you under 419 | this License. If your rights have been terminated and not permanently 420 | reinstated, you do not qualify to receive new licenses for the same 421 | material under section 10. 422 | 423 | 9. Acceptance Not Required for Having Copies. 424 | 425 | You are not required to accept this License in order to receive or 426 | run a copy of the Program. Ancillary propagation of a covered work 427 | occurring solely as a consequence of using peer-to-peer transmission 428 | to receive a copy likewise does not require acceptance. However, 429 | nothing other than this License grants you permission to propagate or 430 | modify any covered work. These actions infringe copyright if you do 431 | not accept this License. Therefore, by modifying or propagating a 432 | covered work, you indicate your acceptance of this License to do so. 433 | 434 | 10. Automatic Licensing of Downstream Recipients. 435 | 436 | Each time you convey a covered work, the recipient automatically 437 | receives a license from the original licensors, to run, modify and 438 | propagate that work, subject to this License. You are not responsible 439 | for enforcing compliance by third parties with this License. 440 | 441 | An "entity transaction" is a transaction transferring control of an 442 | organization, or substantially all assets of one, or subdividing an 443 | organization, or merging organizations. If propagation of a covered 444 | work results from an entity transaction, each party to that 445 | transaction who receives a copy of the work also receives whatever 446 | licenses to the work the party's predecessor in interest had or could 447 | give under the previous paragraph, plus a right to possession of the 448 | Corresponding Source of the work from the predecessor in interest, if 449 | the predecessor has it or can get it with reasonable efforts. 450 | 451 | You may not impose any further restrictions on the exercise of the 452 | rights granted or affirmed under this License. For example, you may 453 | not impose a license fee, royalty, or other charge for exercise of 454 | rights granted under this License, and you may not initiate litigation 455 | (including a cross-claim or counterclaim in a lawsuit) alleging that 456 | any patent claim is infringed by making, using, selling, offering for 457 | sale, or importing the Program or any portion of it. 458 | 459 | 11. Patents. 460 | 461 | A "contributor" is a copyright holder who authorizes use under this 462 | License of the Program or a work on which the Program is based. The 463 | work thus licensed is called the contributor's "contributor version". 464 | 465 | A contributor's "essential patent claims" are all patent claims 466 | owned or controlled by the contributor, whether already acquired or 467 | hereafter acquired, that would be infringed by some manner, permitted 468 | by this License, of making, using, or selling its contributor version, 469 | but do not include claims that would be infringed only as a 470 | consequence of further modification of the contributor version. For 471 | purposes of this definition, "control" includes the right to grant 472 | patent sublicenses in a manner consistent with the requirements of 473 | this License. 474 | 475 | Each contributor grants you a non-exclusive, worldwide, royalty-free 476 | patent license under the contributor's essential patent claims, to 477 | make, use, sell, offer for sale, import and otherwise run, modify and 478 | propagate the contents of its contributor version. 479 | 480 | In the following three paragraphs, a "patent license" is any express 481 | agreement or commitment, however denominated, not to enforce a patent 482 | (such as an express permission to practice a patent or covenant not to 483 | sue for patent infringement). To "grant" such a patent license to a 484 | party means to make such an agreement or commitment not to enforce a 485 | patent against the party. 486 | 487 | If you convey a covered work, knowingly relying on a patent license, 488 | and the Corresponding Source of the work is not available for anyone 489 | to copy, free of charge and under the terms of this License, through a 490 | publicly available network server or other readily accessible means, 491 | then you must either (1) cause the Corresponding Source to be so 492 | available, or (2) arrange to deprive yourself of the benefit of the 493 | patent license for this particular work, or (3) arrange, in a manner 494 | consistent with the requirements of this License, to extend the patent 495 | license to downstream recipients. "Knowingly relying" means you have 496 | actual knowledge that, but for the patent license, your conveying the 497 | covered work in a country, or your recipient's use of the covered work 498 | in a country, would infringe one or more identifiable patents in that 499 | country that you have reason to believe are valid. 500 | 501 | If, pursuant to or in connection with a single transaction or 502 | arrangement, you convey, or propagate by procuring conveyance of, a 503 | covered work, and grant a patent license to some of the parties 504 | receiving the covered work authorizing them to use, propagate, modify 505 | or convey a specific copy of the covered work, then the patent license 506 | you grant is automatically extended to all recipients of the covered 507 | work and works based on it. 508 | 509 | A patent license is "discriminatory" if it does not include within 510 | the scope of its coverage, prohibits the exercise of, or is 511 | conditioned on the non-exercise of one or more of the rights that are 512 | specifically granted under this License. You may not convey a covered 513 | work if you are a party to an arrangement with a third party that is 514 | in the business of distributing software, under which you make payment 515 | to the third party based on the extent of your activity of conveying 516 | the work, and under which the third party grants, to any of the 517 | parties who would receive the covered work from you, a discriminatory 518 | patent license (a) in connection with copies of the covered work 519 | conveyed by you (or copies made from those copies), or (b) primarily 520 | for and in connection with specific products or compilations that 521 | contain the covered work, unless you entered into that arrangement, 522 | or that patent license was granted, prior to 28 March 2007. 523 | 524 | Nothing in this License shall be construed as excluding or limiting 525 | any implied license or other defenses to infringement that may 526 | otherwise be available to you under applicable patent law. 527 | 528 | 12. No Surrender of Others' Freedom. 529 | 530 | If conditions are imposed on you (whether by court order, agreement or 531 | otherwise) that contradict the conditions of this License, they do not 532 | excuse you from the conditions of this License. If you cannot convey a 533 | covered work so as to satisfy simultaneously your obligations under this 534 | License and any other pertinent obligations, then as a consequence you may 535 | not convey it at all. For example, if you agree to terms that obligate you 536 | to collect a royalty for further conveying from those to whom you convey 537 | the Program, the only way you could satisfy both those terms and this 538 | License would be to refrain entirely from conveying the Program. 539 | 540 | 13. Remote Network Interaction; Use with the GNU General Public License. 541 | 542 | Notwithstanding any other provision of this License, if you modify the 543 | Program, your modified version must prominently offer all users 544 | interacting with it remotely through a computer network (if your version 545 | supports such interaction) an opportunity to receive the Corresponding 546 | Source of your version by providing access to the Corresponding Source 547 | from a network server at no charge, through some standard or customary 548 | means of facilitating copying of software. This Corresponding Source 549 | shall include the Corresponding Source for any work covered by version 3 550 | of the GNU General Public License that is incorporated pursuant to the 551 | following paragraph. 552 | 553 | Notwithstanding any other provision of this License, you have 554 | permission to link or combine any covered work with a work licensed 555 | under version 3 of the GNU General Public License into a single 556 | combined work, and to convey the resulting work. The terms of this 557 | License will continue to apply to the part which is the covered work, 558 | but the work with which it is combined will remain governed by version 559 | 3 of the GNU General Public License. 560 | 561 | 14. Revised Versions of this License. 562 | 563 | The Free Software Foundation may publish revised and/or new versions of 564 | the GNU Affero General Public License from time to time. Such new versions 565 | will be similar in spirit to the present version, but may differ in detail to 566 | address new problems or concerns. 567 | 568 | Each version is given a distinguishing version number. If the 569 | Program specifies that a certain numbered version of the GNU Affero General 570 | Public License "or any later version" applies to it, you have the 571 | option of following the terms and conditions either of that numbered 572 | version or of any later version published by the Free Software 573 | Foundation. If the Program does not specify a version number of the 574 | GNU Affero General Public License, you may choose any version ever published 575 | by the Free Software Foundation. 576 | 577 | If the Program specifies that a proxy can decide which future 578 | versions of the GNU Affero General Public License can be used, that proxy's 579 | public statement of acceptance of a version permanently authorizes you 580 | to choose that version for the Program. 581 | 582 | Later license versions may give you additional or different 583 | permissions. However, no additional obligations are imposed on any 584 | author or copyright holder as a result of your choosing to follow a 585 | later version. 586 | 587 | 15. Disclaimer of Warranty. 588 | 589 | THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY 590 | APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT 591 | HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY 592 | OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, 593 | THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 594 | PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM 595 | IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF 596 | ALL NECESSARY SERVICING, REPAIR OR CORRECTION. 597 | 598 | 16. Limitation of Liability. 599 | 600 | IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING 601 | WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MODIFIES AND/OR CONVEYS 602 | THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY 603 | GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE 604 | USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF 605 | DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD 606 | PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS), 607 | EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF 608 | SUCH DAMAGES. 609 | 610 | 17. Interpretation of Sections 15 and 16. 611 | 612 | If the disclaimer of warranty and limitation of liability provided 613 | above cannot be given local legal effect according to their terms, 614 | reviewing courts shall apply local law that most closely approximates 615 | an absolute waiver of all civil liability in connection with the 616 | Program, unless a warranty or assumption of liability accompanies a 617 | copy of the Program in return for a fee. 618 | 619 | END OF TERMS AND CONDITIONS 620 | 621 | How to Apply These Terms to Your New Programs 622 | 623 | If you develop a new program, and you want it to be of the greatest 624 | possible use to the public, the best way to achieve this is to make it 625 | free software which everyone can redistribute and change under these terms. 626 | 627 | To do so, attach the following notices to the program. It is safest 628 | to attach them to the start of each source file to most effectively 629 | state the exclusion of warranty; and each file should have at least 630 | the "copyright" line and a pointer to where the full notice is found. 631 | 632 | 633 | Copyright (C) 634 | 635 | This program is free software: you can redistribute it and/or modify 636 | it under the terms of the GNU Affero General Public License as published 637 | by the Free Software Foundation, either version 3 of the License, or 638 | (at your option) any later version. 639 | 640 | This program is distributed in the hope that it will be useful, 641 | but WITHOUT ANY WARRANTY; without even the implied warranty of 642 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the 643 | GNU Affero General Public License for more details. 644 | 645 | You should have received a copy of the GNU Affero General Public License 646 | along with this program. If not, see . 647 | 648 | Also add information on how to contact you by electronic and paper mail. 649 | 650 | If your software can interact with users remotely through a computer 651 | network, you should also make sure that it provides a way for users to 652 | get its source. For example, if your program is a web application, its 653 | interface could display a "Source" link that leads users to an archive 654 | of the code. There are many ways you could offer source, and different 655 | solutions will be better for different programs; see section 13 for the 656 | specific requirements. 657 | 658 | You should also get your employer (if you work as a programmer) or school, 659 | if any, to sign a "copyright disclaimer" for the program, if necessary. 660 | For more information on this, and how to apply and follow the GNU AGPL, see 661 | . 662 | -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 |

Deep-Live-Cam

2 | 3 |

4 | Real-time face swap and video deepfake with a single click and only a single image. 5 |

6 | 7 |

8 | hacksider%2FDeep-Live-Cam | Trendshift 9 |

10 | 11 |

12 | Demo GIF 13 |

14 | 15 | ## Disclaimer 16 | 17 | This deepfake software is designed to be a productive tool for the AI-generated media industry. It can assist artists in animating custom characters, creating engaging content, and even using models for clothing design. 18 | 19 | We are aware of the potential for unethical applications and are committed to preventative measures. A built-in check prevents the program from processing inappropriate media (nudity, graphic content, sensitive material like war footage, etc.). We will continue to develop this project responsibly, adhering to the law and ethics. We may shut down the project or add watermarks if legally required. 20 | 21 | - Ethical Use: Users are expected to use this software responsibly and legally. If using a real person's face, obtain their consent and clearly label any output as a deepfake when sharing online. 22 | 23 | - Content Restrictions: The software includes built-in checks to prevent processing inappropriate media, such as nudity, graphic content, or sensitive material. 24 | 25 | - Legal Compliance: We adhere to all relevant laws and ethical guidelines. If legally required, we may shut down the project or add watermarks to the output. 26 | 27 | - User Responsibility: We are not responsible for end-user actions. Users must ensure their use of the software aligns with ethical standards and legal requirements. 28 | 29 | By using this software, you agree to these terms and commit to using it in a manner that respects the rights and dignity of others. 30 | 31 | Users are expected to use this software responsibly and legally. If using a real person's face, obtain their consent and clearly label any output as a deepfake when sharing online. We are not responsible for end-user actions. 32 | 33 | ## Exclusive v2.0 Quick Start - Pre-built (Windows) 34 | 35 | 36 | 37 | ##### This is the fastest build you can get if you have a discrete NVIDIA or AMD GPU. 38 | 39 | ###### These Pre-builts are perfect for non-technical users or those who don't have time to, or can't manually install all the requirements. Just a heads-up: this is an open-source project, so you can also install it manually. This will be 60 days ahead on the open source version. 40 | 41 | ## TLDR; Live Deepfake in just 3 Clicks 42 | ![easysteps](https://github.com/user-attachments/assets/af825228-852c-411b-b787-ffd9aac72fc6) 43 | 1. Select a face 44 | 2. Select which camera to use 45 | 3. Press live! 46 | 47 | ## Features & Uses - Everything is in real-time 48 | 49 | ### Mouth Mask 50 | 51 | **Retain your original mouth for accurate movement using Mouth Mask** 52 | 53 |

54 | resizable-gif 55 |

56 | 57 | ### Face Mapping 58 | 59 | **Use different faces on multiple subjects simultaneously** 60 | 61 |

62 | face_mapping_source 63 |

64 | 65 | ### Your Movie, Your Face 66 | 67 | **Watch movies with any face in real-time** 68 | 69 |

70 | movie 71 |

72 | 73 | ### Live Show 74 | 75 | **Run Live shows and performances** 76 | 77 |

78 | show 79 |

80 | 81 | ### Memes 82 | 83 | **Create Your Most Viral Meme Yet** 84 | 85 |

86 | show 87 |
88 | Created using Many Faces feature in Deep-Live-Cam 89 |

90 | 91 | ### Omegle 92 | 93 | **Surprise people on Omegle** 94 | 95 |

96 | 97 |

98 | 99 | ## Installation (Manual) 100 | 101 | **Please be aware that the installation requires technical skills and is not for beginners. Consider downloading the prebuilt version.** 102 | 103 |
104 | Click to see the process 105 | 106 | ### Installation 107 | 108 | This is more likely to work on your computer but will be slower as it utilizes the CPU. 109 | 110 | **1. Set up Your Platform** 111 | 112 | - Python (3.10 recommended) 113 | - pip 114 | - git 115 | - [ffmpeg](https://www.youtube.com/watch?v=OlNWCpFdVMA) - ```iex (irm ffmpeg.tc.ht)``` 116 | - [Visual Studio 2022 Runtimes (Windows)](https://visualstudio.microsoft.com/visual-cpp-build-tools/) 117 | 118 | **2. Clone the Repository** 119 | 120 | ```bash 121 | git clone https://github.com/hacksider/Deep-Live-Cam.git 122 | cd Deep-Live-Cam 123 | ``` 124 | 125 | **3. Download the Models** 126 | 127 | 1. [GFPGANv1.4](https://huggingface.co/hacksider/deep-live-cam/resolve/main/GFPGANv1.4.pth) 128 | 2. [inswapper\_128\_fp16.onnx](https://huggingface.co/hacksider/deep-live-cam/resolve/main/inswapper_128_fp16.onnx) 129 | 130 | Place these files in the "**models**" folder. 131 | 132 | **4. Install Dependencies** 133 | 134 | We highly recommend using a `venv` to avoid issues. 135 | 136 | 137 | For Windows: 138 | ```bash 139 | python -m venv venv 140 | venv\Scripts\activate 141 | pip install -r requirements.txt 142 | ``` 143 | For Linux: 144 | ```bash 145 | # Ensure you use the installed Python 3.10 146 | python3 -m venv venv 147 | source venv/bin/activate 148 | pip install -r requirements.txt 149 | ``` 150 | 151 | **For macOS:** 152 | 153 | Apple Silicon (M1/M2/M3) requires specific setup: 154 | 155 | ```bash 156 | # Install Python 3.10 (specific version is important) 157 | brew install python@3.10 158 | 159 | # Install tkinter package (required for the GUI) 160 | brew install python-tk@3.10 161 | 162 | # Create and activate virtual environment with Python 3.10 163 | python3.10 -m venv venv 164 | source venv/bin/activate 165 | 166 | # Install dependencies 167 | pip install -r requirements.txt 168 | ``` 169 | 170 | ** In case something goes wrong and you need to reinstall the virtual environment ** 171 | 172 | ```bash 173 | # Deactivate the virtual environment 174 | rm -rf venv 175 | 176 | # Reinstall the virtual environment 177 | python -m venv venv 178 | source venv/bin/activate 179 | 180 | # install the dependencies again 181 | pip install -r requirements.txt 182 | ``` 183 | 184 | **Run:** If you don't have a GPU, you can run Deep-Live-Cam using `python run.py`. Note that initial execution will download models (~300MB). 185 | 186 | ### GPU Acceleration 187 | 188 | **CUDA Execution Provider (Nvidia)** 189 | 190 | 1. Install [CUDA Toolkit 11.8.0](https://developer.nvidia.com/cuda-11-8-0-download-archive) 191 | 2. Install [cuDNN v8.9.7 for CUDA 11.x](https://developer.nvidia.com/rdp/cudnn-archive) (required for onnxruntime-gpu): 192 | - Download cuDNN v8.9.7 for CUDA 11.x 193 | - Make sure the cuDNN bin directory is in your system PATH 194 | 3. Install dependencies: 195 | 196 | ```bash 197 | pip uninstall onnxruntime onnxruntime-gpu 198 | pip install onnxruntime-gpu==1.16.3 199 | ``` 200 | 201 | 3. Usage: 202 | 203 | ```bash 204 | python run.py --execution-provider cuda 205 | ``` 206 | 207 | **CoreML Execution Provider (Apple Silicon)** 208 | 209 | Apple Silicon (M1/M2/M3) specific installation: 210 | 211 | 1. Make sure you've completed the macOS setup above using Python 3.10. 212 | 2. Install dependencies: 213 | 214 | ```bash 215 | pip uninstall onnxruntime onnxruntime-silicon 216 | pip install onnxruntime-silicon==1.13.1 217 | ``` 218 | 219 | 3. Usage (important: specify Python 3.10): 220 | 221 | ```bash 222 | python3.10 run.py --execution-provider coreml 223 | ``` 224 | 225 | **Important Notes for macOS:** 226 | - You **must** use Python 3.10, not newer versions like 3.11 or 3.13 227 | - Always run with `python3.10` command not just `python` if you have multiple Python versions installed 228 | - If you get error about `_tkinter` missing, reinstall the tkinter package: `brew reinstall python-tk@3.10` 229 | - If you get model loading errors, check that your models are in the correct folder 230 | - If you encounter conflicts with other Python versions, consider uninstalling them: 231 | ```bash 232 | # List all installed Python versions 233 | brew list | grep python 234 | 235 | # Uninstall conflicting versions if needed 236 | brew uninstall --ignore-dependencies python@3.11 python@3.13 237 | 238 | # Keep only Python 3.10 239 | brew cleanup 240 | ``` 241 | 242 | **CoreML Execution Provider (Apple Legacy)** 243 | 244 | 1. Install dependencies: 245 | 246 | ```bash 247 | pip uninstall onnxruntime onnxruntime-coreml 248 | pip install onnxruntime-coreml==1.13.1 249 | ``` 250 | 251 | 2. Usage: 252 | 253 | ```bash 254 | python run.py --execution-provider coreml 255 | ``` 256 | 257 | **DirectML Execution Provider (Windows)** 258 | 259 | 1. Install dependencies: 260 | 261 | ```bash 262 | pip uninstall onnxruntime onnxruntime-directml 263 | pip install onnxruntime-directml==1.15.1 264 | ``` 265 | 266 | 2. Usage: 267 | 268 | ```bash 269 | python run.py --execution-provider directml 270 | ``` 271 | 272 | **OpenVINO™ Execution Provider (Intel)** 273 | 274 | 1. Install dependencies: 275 | 276 | ```bash 277 | pip uninstall onnxruntime onnxruntime-openvino 278 | pip install onnxruntime-openvino==1.15.0 279 | ``` 280 | 281 | 2. Usage: 282 | 283 | ```bash 284 | python run.py --execution-provider openvino 285 | ``` 286 |
287 | 288 | ## Usage 289 | 290 | **1. Image/Video Mode** 291 | 292 | - Execute `python run.py`. 293 | - Choose a source face image and a target image/video. 294 | - Click "Start". 295 | - The output will be saved in a directory named after the target video. 296 | 297 | **2. Webcam Mode** 298 | 299 | - Execute `python run.py`. 300 | - Select a source face image. 301 | - Click "Live". 302 | - Wait for the preview to appear (10-30 seconds). 303 | - Use a screen capture tool like OBS to stream. 304 | - To change the face, select a new source image. 305 | 306 | ## Tips and Tricks 307 | 308 | Check out these helpful guides to get the most out of Deep-Live-Cam: 309 | 310 | - [Unlocking the Secrets to the Perfect Deepfake Image](https://deeplivecam.net/index.php/blog/tips-and-tricks/unlocking-the-secrets-to-the-perfect-deepfake-image) - Learn how to create the best deepfake with full head coverage 311 | - [Video Call with DeepLiveCam](https://deeplivecam.net/index.php/blog/tips-and-tricks/video-call-with-deeplivecam) - Make your meetings livelier by using DeepLiveCam with OBS and meeting software 312 | - [Have a Special Guest!](https://deeplivecam.net/index.php/blog/tips-and-tricks/have-a-special-guest) - Tutorial on how to use face mapping to add special guests to your stream 313 | - [Watch Deepfake Movies in Realtime](https://deeplivecam.net/index.php/blog/tips-and-tricks/watch-deepfake-movies-in-realtime) - See yourself star in any video without processing the video 314 | - [Better Quality without Sacrificing Speed](https://deeplivecam.net/index.php/blog/tips-and-tricks/better-quality-without-sacrificing-speed) - Tips for achieving better results without impacting performance 315 | - [Instant Vtuber!](https://deeplivecam.net/index.php/blog/tips-and-tricks/instant-vtuber) - Create a new persona/vtuber easily using Metahuman Creator 316 | 317 | Visit our [official blog](https://deeplivecam.net/index.php/blog/tips-and-tricks) for more tips and tutorials. 318 | 319 | ## Command Line Arguments (Unmaintained) 320 | 321 | ``` 322 | options: 323 | -h, --help show this help message and exit 324 | -s SOURCE_PATH, --source SOURCE_PATH select a source image 325 | -t TARGET_PATH, --target TARGET_PATH select a target image or video 326 | -o OUTPUT_PATH, --output OUTPUT_PATH select output file or directory 327 | --frame-processor FRAME_PROCESSOR [FRAME_PROCESSOR ...] frame processors (choices: face_swapper, face_enhancer, ...) 328 | --keep-fps keep original fps 329 | --keep-audio keep original audio 330 | --keep-frames keep temporary frames 331 | --many-faces process every face 332 | --map-faces map source target faces 333 | --mouth-mask mask the mouth region 334 | --video-encoder {libx264,libx265,libvpx-vp9} adjust output video encoder 335 | --video-quality [0-51] adjust output video quality 336 | --live-mirror the live camera display as you see it in the front-facing camera frame 337 | --live-resizable the live camera frame is resizable 338 | --max-memory MAX_MEMORY maximum amount of RAM in GB 339 | --execution-provider {cpu} [{cpu} ...] available execution provider (choices: cpu, ...) 340 | --execution-threads EXECUTION_THREADS number of execution threads 341 | -v, --version show program's version number and exit 342 | ``` 343 | 344 | Looking for a CLI mode? Using the -s/--source argument will make the run program in cli mode. 345 | 346 | ## Press 347 | 348 | **We are always open to criticism and are ready to improve, that's why we didn't cherry-pick anything.** 349 | 350 | - [*"Deep-Live-Cam goes viral, allowing anyone to become a digital doppelganger"*](https://arstechnica.com/information-technology/2024/08/new-ai-tool-enables-real-time-face-swapping-on-webcams-raising-fraud-concerns/) - Ars Technica 351 | - [*"Thanks Deep Live Cam, shapeshifters are among us now"*](https://dataconomy.com/2024/08/15/what-is-deep-live-cam-github-deepfake/) - Dataconomy 352 | - [*"This free AI tool lets you become anyone during video-calls"*](https://www.newsbytesapp.com/news/science/deep-live-cam-ai-impersonation-tool-goes-viral/story) - NewsBytes 353 | - [*"OK, this viral AI live stream software is truly terrifying"*](https://www.creativebloq.com/ai/ok-this-viral-ai-live-stream-software-is-truly-terrifying) - Creative Bloq 354 | - [*"Deepfake AI Tool Lets You Become Anyone in a Video Call With Single Photo"*](https://petapixel.com/2024/08/14/deep-live-cam-deepfake-ai-tool-lets-you-become-anyone-in-a-video-call-with-single-photo-mark-zuckerberg-jd-vance-elon-musk/) - PetaPixel 355 | - [*"Deep-Live-Cam Uses AI to Transform Your Face in Real-Time, Celebrities Included"*](https://www.techeblog.com/deep-live-cam-ai-transform-face/) - TechEBlog 356 | - [*"An AI tool that "makes you look like anyone" during a video call is going viral online"*](https://telegrafi.com/en/a-tool-that-makes-you-look-like-anyone-during-a-video-call-is-going-viral-on-the-Internet/) - Telegrafi 357 | - [*"This Deepfake Tool Turning Images Into Livestreams is Topping the GitHub Charts"*](https://decrypt.co/244565/this-deepfake-tool-turning-images-into-livestreams-is-topping-the-github-charts) - Emerge 358 | - [*"New Real-Time Face-Swapping AI Allows Anyone to Mimic Famous Faces"*](https://www.digitalmusicnews.com/2024/08/15/face-swapping-ai-real-time-mimic/) - Digital Music News 359 | - [*"This real-time webcam deepfake tool raises alarms about the future of identity theft"*](https://www.diyphotography.net/this-real-time-webcam-deepfake-tool-raises-alarms-about-the-future-of-identity-theft/) - DIYPhotography 360 | - [*"That's Crazy, Oh God. That's Fucking Freaky Dude... That's So Wild Dude"*](https://www.youtube.com/watch?time_continue=1074&v=py4Tc-Y8BcY) - SomeOrdinaryGamers 361 | - [*"Alright look look look, now look chat, we can do any face we want to look like chat"*](https://www.youtube.com/live/mFsCe7AIxq8?feature=shared&t=2686) - IShowSpeed 362 | 363 | ## Credits 364 | 365 | - [ffmpeg](https://ffmpeg.org/): for making video-related operations easy 366 | - [deepinsight](https://github.com/deepinsight): for their [insightface](https://github.com/deepinsight/insightface) project which provided a well-made library and models. Please be reminded that the [use of the model is for non-commercial research purposes only](https://github.com/deepinsight/insightface?tab=readme-ov-file#license). 367 | - [havok2-htwo](https://github.com/havok2-htwo): for sharing the code for webcam 368 | - [GosuDRM](https://github.com/GosuDRM): for the open version of roop 369 | - [pereiraroland26](https://github.com/pereiraroland26): Multiple faces support 370 | - [vic4key](https://github.com/vic4key): For supporting/contributing to this project 371 | - [kier007](https://github.com/kier007): for improving the user experience 372 | - [qitianai](https://github.com/qitianai): for multi-lingual support 373 | - and [all developers](https://github.com/hacksider/Deep-Live-Cam/graphs/contributors) behind libraries used in this project. 374 | - Footnote: Please be informed that the base author of the code is [s0md3v](https://github.com/s0md3v/roop) 375 | - All the wonderful users who helped make this project go viral by starring the repo ❤️ 376 | 377 | [![Stargazers](https://reporoster.com/stars/hacksider/Deep-Live-Cam)](https://github.com/hacksider/Deep-Live-Cam/stargazers) 378 | 379 | ## Contributions 380 | 381 | ![Alt](https://repobeats.axiom.co/api/embed/fec8e29c45dfdb9c5916f3a7830e1249308d20e1.svg "Repobeats analytics image") 382 | 383 | ## Stars to the Moon 🚀 384 | 385 | 386 | 387 | 388 | 389 | Star History Chart 390 | 391 | 392 | -------------------------------------------------------------------------------- /locales/de.json: -------------------------------------------------------------------------------- 1 | { 2 | "Source x Target Mapper": "Quelle x Ziel Zuordnung", 3 | "select a source image": "Wähle ein Quellbild", 4 | "Preview": "Vorschau", 5 | "select a target image or video": "Wähle ein Zielbild oder Video", 6 | "save image output file": "Bildausgabedatei speichern", 7 | "save video output file": "Videoausgabedatei speichern", 8 | "select a target image": "Wähle ein Zielbild", 9 | "source": "Quelle", 10 | "Select a target": "Wähle ein Ziel", 11 | "Select a face": "Wähle ein Gesicht", 12 | "Keep audio": "Audio beibehalten", 13 | "Face Enhancer": "Gesichtsverbesserung", 14 | "Many faces": "Mehrere Gesichter", 15 | "Show FPS": "FPS anzeigen", 16 | "Keep fps": "FPS beibehalten", 17 | "Keep frames": "Frames beibehalten", 18 | "Fix Blueish Cam": "Bläuliche Kamera korrigieren", 19 | "Mouth Mask": "Mundmaske", 20 | "Show Mouth Mask Box": "Mundmaskenrahmen anzeigen", 21 | "Start": "Starten", 22 | "Live": "Live", 23 | "Destroy": "Beenden", 24 | "Map faces": "Gesichter zuordnen", 25 | "Processing...": "Verarbeitung läuft...", 26 | "Processing succeed!": "Verarbeitung erfolgreich!", 27 | "Processing ignored!": "Verarbeitung ignoriert!", 28 | "Failed to start camera": "Kamera konnte nicht gestartet werden", 29 | "Please complete pop-up or close it.": "Bitte das Pop-up komplettieren oder schließen.", 30 | "Getting unique faces": "Einzigartige Gesichter erfassen", 31 | "Please select a source image first": "Bitte zuerst ein Quellbild auswählen", 32 | "No faces found in target": "Keine Gesichter im Zielbild gefunden", 33 | "Add": "Hinzufügen", 34 | "Clear": "Löschen", 35 | "Submit": "Absenden", 36 | "Select source image": "Quellbild auswählen", 37 | "Select target image": "Zielbild auswählen", 38 | "Please provide mapping!": "Bitte eine Zuordnung angeben!", 39 | "At least 1 source with target is required!": "Mindestens eine Quelle mit einem Ziel ist erforderlich!", 40 | "At least 1 source with target is required!": "Mindestens eine Quelle mit einem Ziel ist erforderlich!", 41 | "Face could not be detected in last upload!": "Im letzten Upload konnte kein Gesicht erkannt werden!", 42 | "Select Camera:": "Kamera auswählen:", 43 | "All mappings cleared!": "Alle Zuordnungen gelöscht!", 44 | "Mappings successfully submitted!": "Zuordnungen erfolgreich übermittelt!", 45 | "Source x Target Mapper is already open.": "Quell-zu-Ziel-Zuordnung ist bereits geöffnet." 46 | } 47 | -------------------------------------------------------------------------------- /locales/es.json: -------------------------------------------------------------------------------- 1 | { 2 | "Source x Target Mapper": "Mapeador de fuente x destino", 3 | "select a source image": "Seleccionar imagen fuente", 4 | "Preview": "Vista previa", 5 | "select a target image or video": "elegir un video o una imagen fuente", 6 | "save image output file": "guardar imagen final", 7 | "save video output file": "guardar video final", 8 | "select a target image": "elegir una imagen objetiva", 9 | "source": "fuente", 10 | "Select a target": "Elegir un destino", 11 | "Select a face": "Elegir una cara", 12 | "Keep audio": "Mantener audio original", 13 | "Face Enhancer": "Potenciador de caras", 14 | "Many faces": "Varias caras", 15 | "Show FPS": "Mostrar fps", 16 | "Keep fps": "Mantener fps", 17 | "Keep frames": "Mantener frames", 18 | "Fix Blueish Cam": "Corregir tono azul de video", 19 | "Mouth Mask": "Máscara de boca", 20 | "Show Mouth Mask Box": "Mostrar área de la máscara de boca", 21 | "Start": "Iniciar", 22 | "Live": "En vivo", 23 | "Destroy": "Borrar", 24 | "Map faces": "Mapear caras", 25 | "Processing...": "Procesando...", 26 | "Processing succeed!": "¡Proceso terminado con éxito!", 27 | "Processing ignored!": "¡Procesamiento omitido!", 28 | "Failed to start camera": "No se pudo iniciar la cámara", 29 | "Please complete pop-up or close it.": "Complete o cierre el pop-up", 30 | "Getting unique faces": "Buscando caras únicas", 31 | "Please select a source image first": "Primero, seleccione una imagen fuente", 32 | "No faces found in target": "No se encontró una cara en el destino", 33 | "Add": "Agregar", 34 | "Clear": "Limpiar", 35 | "Submit": "Enviar", 36 | "Select source image": "Seleccionar imagen fuente", 37 | "Select target image": "Seleccionar imagen destino", 38 | "Please provide mapping!": "Por favor, proporcione un mapeo", 39 | "At least 1 source with target is required!": "Se requiere al menos una fuente con un destino.", 40 | "At least 1 source with target is required!": "Se requiere al menos una fuente con un destino.", 41 | "Face could not be detected in last upload!": "¡No se pudo encontrar una cara en el último video o imagen!", 42 | "Select Camera:": "Elegir cámara:", 43 | "All mappings cleared!": "¡Todos los mapeos fueron borrados!", 44 | "Mappings successfully submitted!": "Mapeos enviados con éxito!", 45 | "Source x Target Mapper is already open.": "El mapeador de fuente x destino ya está abierto." 46 | } -------------------------------------------------------------------------------- /locales/fi.json: -------------------------------------------------------------------------------- 1 | { 2 | "Source x Target Mapper": "Source x Target Kartoitin", 3 | "select an source image": "Valitse lähde kuva", 4 | "Preview": "Esikatsele", 5 | "select an target image or video": "Valitse kohde kuva tai video", 6 | "save image output file": "tallenna kuva", 7 | "save video output file": "tallenna video", 8 | "select an target image": "Valitse kohde kuva", 9 | "source": "lähde", 10 | "Select a target": "Valitse kohde", 11 | "Select a face": "Valitse kasvot", 12 | "Keep audio": "Säilytä ääni", 13 | "Face Enhancer": "Kasvojen Parantaja", 14 | "Many faces": "Useampia kasvoja", 15 | "Show FPS": "Näytä FPS", 16 | "Keep fps": "Säilytä FPS", 17 | "Keep frames": "Säilytä ruudut", 18 | "Fix Blueish Cam": "Korjaa Sinertävä Kamera", 19 | "Mouth Mask": "Suu Maski", 20 | "Show Mouth Mask Box": "Näytä Suu Maski Laatiko", 21 | "Start": "Aloita", 22 | "Live": "Live", 23 | "Destroy": "Tuhoa", 24 | "Map faces": "Kartoita kasvot", 25 | "Processing...": "Prosessoi...", 26 | "Processing succeed!": "Prosessointi onnistui!", 27 | "Processing ignored!": "Prosessointi lopetettu!", 28 | "Failed to start camera": "Kameran käynnistäminen epäonnistui", 29 | "Please complete pop-up or close it.": "Viimeistele tai sulje ponnahdusikkuna", 30 | "Getting unique faces": "Hankitaan uniikkeja kasvoja", 31 | "Please select a source image first": "Valitse ensin lähde kuva", 32 | "No faces found in target": "Kasvoja ei löydetty kohteessa", 33 | "Add": "Lisää", 34 | "Clear": "Tyhjennä", 35 | "Submit": "Lähetä", 36 | "Select source image": "Valitse lähde kuva", 37 | "Select target image": "Valitse kohde kuva", 38 | "Please provide mapping!": "Tarjoa kartoitus!", 39 | "Atleast 1 source with target is required!": "Vähintään 1 lähde kohteen kanssa on vaadittu!", 40 | "At least 1 source with target is required!": "Vähintään 1 lähde kohteen kanssa on vaadittu!", 41 | "Face could not be detected in last upload!": "Kasvoja ei voitu tunnistaa edellisessä latauksessa!", 42 | "Select Camera:": "Valitse Kamera:", 43 | "All mappings cleared!": "Kaikki kartoitukset tyhjennetty!", 44 | "Mappings successfully submitted!": "Kartoitukset lähetety onnistuneesti!", 45 | "Source x Target Mapper is already open.": "Lähde x Kohde Kartoittaja on jo auki." 46 | } 47 | -------------------------------------------------------------------------------- /locales/km.json: -------------------------------------------------------------------------------- 1 | { 2 | "Source x Target Mapper": "ប្រភប x បន្ថែម Mapper", 3 | "select a source image": "ជ្រើសរើសប្រភពរូបភាព", 4 | "Preview": "បង្ហាញ", 5 | "select a target image or video": "ជ្រើសរើសគោលដៅរូបភាពឬវីដេអូ", 6 | "save image output file": "រក្សាទុកលទ្ធផលឯកសាររូបភាព", 7 | "save video output file": "រក្សាទុកលទ្ធផលឯកសារវីដេអូ", 8 | "select a target image": "ជ្រើសរើសគោលដៅរូបភាព", 9 | "source": "ប្រភព", 10 | "Select a target": "ជ្រើសរើសគោលដៅ", 11 | "Select a face": "ជ្រើសរើសមុខ", 12 | "Keep audio": "រម្លងសម្លេង", 13 | "Face Enhancer": "ឧបករណ៍ពង្រឹងមុខ", 14 | "Many faces": "ទម្រង់មុខច្រើន", 15 | "Show FPS": "បង្ហាញ FPS", 16 | "Keep fps": "រម្លង fps", 17 | "Keep frames": "រម្លងទម្រង់", 18 | "Fix Blueish Cam": "ជួសជុល Cam Blueish", 19 | "Mouth Mask": "របាំងមាត់", 20 | "Show Mouth Mask Box": "បង្ហាញប្រអប់របាំងមាត់", 21 | "Start": "ចាប់ផ្ដើម", 22 | "Live": "ផ្សាយផ្ទាល់", 23 | "Destroy": "លុប", 24 | "Map faces": "ផែនទីមុខ", 25 | "Processing...": "កំពុងដំណើរការ...", 26 | "Processing succeed!": "ការដំណើរការទទួលបានជោគជ័យ!", 27 | "Processing ignored!": "ការដំណើរការមិនទទួលបានជោគជ័យ!", 28 | "Failed to start camera": "បរាជ័យដើម្បីចាប់ផ្ដើមបើកកាមេរ៉ា", 29 | "Please complete pop-up or close it.": "សូមបញ្ចប់ផ្ទាំងផុស ឬបិទវា.", 30 | "Getting unique faces": "ការចាប់ផ្ដើមទម្រង់មុខប្លែក", 31 | "Please select a source image first": "សូមជ្រើសរើសប្រភពរូបភាពដំបូង", 32 | "No faces found in target": "រកអត់ឃើញមុខនៅក្នុងគោលដៅ", 33 | "Add": "បន្ថែម", 34 | "Clear": "សម្អាត", 35 | "Submit": "បញ្ចូន", 36 | "Select source image": "ជ្រើសរើសប្រភពរូបភាព", 37 | "Select target image": "ជ្រើសរើសគោលដៅរូបភាព", 38 | "Please provide mapping!": "សូមផ្ដល់នៅផែនទី", 39 | "At least 1 source with target is required!": "ត្រូវការប្រភពយ៉ាងហោចណាស់ ១ ដែលមានគោលដៅ!", 40 | "Face could not be detected in last upload!": "មុខមិនអាចភ្ជាប់នៅក្នុងការបង្ហេាះចុងក្រោយ!", 41 | "Select Camera:": "ជ្រើសរើសកាមេរ៉ា", 42 | "All mappings cleared!": "ផែនទីទាំងអស់ត្រូវបានសម្អាត!", 43 | "Mappings successfully submitted!": "ផែនទីត្រូវបានបញ្ជូនជោគជ័យ!", 44 | "Source x Target Mapper is already open.": "ប្រភព x Target Mapper បានបើករួចហើយ។" 45 | } 46 | -------------------------------------------------------------------------------- /locales/ko.json: -------------------------------------------------------------------------------- 1 | { 2 | "Source x Target Mapper": "소스 x 타겟 매퍼", 3 | "select a source image": "소스 이미지 선택", 4 | "Preview": "미리보기", 5 | "select a target image or video": "타겟 이미지 또는 영상 선택", 6 | "save image output file": "이미지 출력 파일 저장", 7 | "save video output file": "영상 출력 파일 저장", 8 | "select a target image": "타겟 이미지 선택", 9 | "source": "소스", 10 | "Select a target": "타겟 선택", 11 | "Select a face": "얼굴 선택", 12 | "Keep audio": "오디오 유지", 13 | "Face Enhancer": "얼굴 향상", 14 | "Many faces": "여러 얼굴", 15 | "Show FPS": "FPS 표시", 16 | "Keep fps": "FPS 유지", 17 | "Keep frames": "프레임 유지", 18 | "Fix Blueish Cam": "푸른빛 카메라 보정", 19 | "Mouth Mask": "입 마스크", 20 | "Show Mouth Mask Box": "입 마스크 박스 표시", 21 | "Start": "시작", 22 | "Live": "라이브", 23 | "Destroy": "종료", 24 | "Map faces": "얼굴 매핑", 25 | "Processing...": "처리 중...", 26 | "Processing succeed!": "처리 성공!", 27 | "Processing ignored!": "처리 무시됨!", 28 | "Failed to start camera": "카메라 시작 실패", 29 | "Please complete pop-up or close it.": "팝업을 완료하거나 닫아주세요.", 30 | "Getting unique faces": "고유 얼굴 가져오는 중", 31 | "Please select a source image first": "먼저 소스 이미지를 선택해주세요", 32 | "No faces found in target": "타겟에서 얼굴을 찾을 수 없음", 33 | "Add": "추가", 34 | "Clear": "지우기", 35 | "Submit": "제출", 36 | "Select source image": "소스 이미지 선택", 37 | "Select target image": "타겟 이미지 선택", 38 | "Please provide mapping!": "매핑을 입력해주세요!", 39 | "At least 1 source with target is required!": "최소 하나의 소스와 타겟이 필요합니다!", 40 | "Face could not be detected in last upload!": "최근 업로드에서 얼굴을 감지할 수 없습니다!", 41 | "Select Camera:": "카메라 선택:", 42 | "All mappings cleared!": "모든 매핑이 삭제되었습니다!", 43 | "Mappings successfully submitted!": "매핑이 성공적으로 제출되었습니다!", 44 | "Source x Target Mapper is already open.": "소스 x 타겟 매퍼가 이미 열려 있습니다." 45 | } 46 | -------------------------------------------------------------------------------- /locales/pt-br.json: -------------------------------------------------------------------------------- 1 | { 2 | "Source x Target Mapper": "Mapeador de Origem x Destino", 3 | "select an source image": "Escolha uma imagem de origem", 4 | "Preview": "Prévia", 5 | "select an target image or video": "Escolha uma imagem ou vídeo de destino", 6 | "save image output file": "Salvar imagem final", 7 | "save video output file": "Salvar vídeo final", 8 | "select an target image": "Escolha uma imagem de destino", 9 | "source": "Origem", 10 | "Select a target": "Escolha o destino", 11 | "Select a face": "Escolha um rosto", 12 | "Keep audio": "Manter o áudio original", 13 | "Face Enhancer": "Melhorar rosto", 14 | "Many faces": "Vários rostos", 15 | "Show FPS": "Mostrar FPS", 16 | "Keep fps": "Manter FPS", 17 | "Keep frames": "Manter frames", 18 | "Fix Blueish Cam": "Corrigir tom azulado da câmera", 19 | "Mouth Mask": "Máscara da boca", 20 | "Show Mouth Mask Box": "Mostrar área da máscara da boca", 21 | "Start": "Começar", 22 | "Live": "Ao vivo", 23 | "Destroy": "Destruir", 24 | "Map faces": "Mapear rostos", 25 | "Processing...": "Processando...", 26 | "Processing succeed!": "Tudo certo!", 27 | "Processing ignored!": "Processamento ignorado!", 28 | "Failed to start camera": "Não foi possível iniciar a câmera", 29 | "Please complete pop-up or close it.": "Finalize ou feche o pop-up", 30 | "Getting unique faces": "Buscando rostos diferentes", 31 | "Please select a source image first": "Selecione primeiro uma imagem de origem", 32 | "No faces found in target": "Nenhum rosto encontrado na imagem de destino", 33 | "Add": "Adicionar", 34 | "Clear": "Limpar", 35 | "Submit": "Enviar", 36 | "Select source image": "Escolha a imagem de origem", 37 | "Select target image": "Escolha a imagem de destino", 38 | "Please provide mapping!": "Você precisa realizar o mapeamento!", 39 | "Atleast 1 source with target is required!": "É necessária pelo menos uma origem com um destino!", 40 | "At least 1 source with target is required!": "É necessária pelo menos uma origem com um destino!", 41 | "Face could not be detected in last upload!": "Não conseguimos detectar o rosto na última imagem!", 42 | "Select Camera:": "Escolher câmera:", 43 | "All mappings cleared!": "Todos os mapeamentos foram removidos!", 44 | "Mappings successfully submitted!": "Mapeamentos enviados com sucesso!", 45 | "Source x Target Mapper is already open.": "O Mapeador de Origem x Destino já está aberto." 46 | } 47 | -------------------------------------------------------------------------------- /locales/ru.json: -------------------------------------------------------------------------------- 1 | { 2 | "Source x Target Mapper": "Сопоставитель Источник x Цель", 3 | "select a source image": "выберите исходное изображение", 4 | "Preview": "Предпросмотр", 5 | "select a target image or video": "выберите целевое изображение или видео", 6 | "save image output file": "сохранить выходной файл изображения", 7 | "save video output file": "сохранить выходной файл видео", 8 | "select a target image": "выберите целевое изображение", 9 | "source": "источник", 10 | "Select a target": "Выберите целевое изображение", 11 | "Select a face": "Выберите лицо", 12 | "Keep audio": "Сохранить аудио", 13 | "Face Enhancer": "Улучшение лица", 14 | "Many faces": "Несколько лиц", 15 | "Show FPS": "Показать FPS", 16 | "Keep fps": "Сохранить FPS", 17 | "Keep frames": "Сохранить кадры", 18 | "Fix Blueish Cam": "Исправить синеву камеры", 19 | "Mouth Mask": "Маска рта", 20 | "Show Mouth Mask Box": "Показать рамку маски рта", 21 | "Start": "Старт", 22 | "Live": "В реальном времени", 23 | "Destroy": "Остановить", 24 | "Map faces": "Сопоставить лица", 25 | "Processing...": "Обработка...", 26 | "Processing succeed!": "Обработка успешна!", 27 | "Processing ignored!": "Обработка проигнорирована!", 28 | "Failed to start camera": "Не удалось запустить камеру", 29 | "Please complete pop-up or close it.": "Пожалуйста, заполните всплывающее окно или закройте его.", 30 | "Getting unique faces": "Получение уникальных лиц", 31 | "Please select a source image first": "Сначала выберите исходное изображение, пожалуйста", 32 | "No faces found in target": "В целевом изображении не найдено лиц", 33 | "Add": "Добавить", 34 | "Clear": "Очистить", 35 | "Submit": "Отправить", 36 | "Select source image": "Выбрать исходное изображение", 37 | "Select target image": "Выбрать целевое изображение", 38 | "Please provide mapping!": "Пожалуйста, укажите сопоставление!", 39 | "At least 1 source with target is required!": "Требуется хотя бы 1 источник с целью!", 40 | "Face could not be detected in last upload!": "Лицо не обнаружено в последнем загруженном изображении!", 41 | "Select Camera:": "Выберите камеру:", 42 | "All mappings cleared!": "Все сопоставления очищены!", 43 | "Mappings successfully submitted!": "Сопоставления успешно отправлены!", 44 | "Source x Target Mapper is already open.": "Сопоставитель Источник-Цель уже открыт." 45 | } -------------------------------------------------------------------------------- /locales/th.json: -------------------------------------------------------------------------------- 1 | { 2 | "Source x Target Mapper": "ตัวจับคู่ต้นทาง x ปลายทาง", 3 | "select a source image": "เลือกรูปภาพต้นฉบับ", 4 | "Preview": "ตัวอย่าง", 5 | "select a target image or video": "เลือกรูปภาพหรือวิดีโอเป้าหมาย", 6 | "save image output file": "บันทึกไฟล์รูปภาพ", 7 | "save video output file": "บันทึกไฟล์วิดีโอ", 8 | "select a target image": "เลือกรูปภาพเป้าหมาย", 9 | "source": "ต้นฉบับ", 10 | "Select a target": "เลือกเป้าหมาย", 11 | "Select a face": "เลือกใบหน้า", 12 | "Keep audio": "เก็บเสียง", 13 | "Face Enhancer": "ปรับปรุงใบหน้า", 14 | "Many faces": "หลายใบหน้า", 15 | "Show FPS": "แสดง FPS", 16 | "Keep fps": "คงค่า FPS", 17 | "Keep frames": "คงค่าเฟรม", 18 | "Fix Blueish Cam": "แก้ไขภาพอมฟ้าจากกล้อง", 19 | "Mouth Mask": "มาสก์ปาก", 20 | "Show Mouth Mask Box": "แสดงกรอบมาสก์ปาก", 21 | "Start": "เริ่ม", 22 | "Live": "สด", 23 | "Destroy": "หยุด", 24 | "Map faces": "จับคู่ใบหน้า", 25 | "Processing...": "กำลังประมวลผล...", 26 | "Processing succeed!": "ประมวลผลสำเร็จแล้ว!", 27 | "Processing ignored!": "การประมวลผลถูกละเว้น", 28 | "Failed to start camera": "ไม่สามารถเริ่มกล้องได้", 29 | "Please complete pop-up or close it.": "โปรดดำเนินการในป๊อปอัปให้เสร็จสิ้น หรือปิด", 30 | "Getting unique faces": "กำลังค้นหาใบหน้าที่ไม่ซ้ำกัน", 31 | "Please select a source image first": "โปรดเลือกภาพต้นฉบับก่อน", 32 | "No faces found in target": "ไม่พบใบหน้าในภาพเป้าหมาย", 33 | "Add": "เพิ่ม", 34 | "Clear": "ล้าง", 35 | "Submit": "ส่ง", 36 | "Select source image": "เลือกภาพต้นฉบับ", 37 | "Select target image": "เลือกภาพเป้าหมาย", 38 | "Please provide mapping!": "โปรดระบุการจับคู่!", 39 | "At least 1 source with target is required!": "ต้องมีการจับคู่ต้นฉบับกับเป้าหมายอย่างน้อย 1 คู่!", 40 | "Face could not be detected in last upload!": "ไม่สามารถตรวจพบใบหน้าในไฟล์อัปโหลดล่าสุด!", 41 | "Select Camera:": "เลือกกล้อง:", 42 | "All mappings cleared!": "ล้างการจับคู่ทั้งหมดแล้ว!", 43 | "Mappings successfully submitted!": "ส่งการจับคู่สำเร็จแล้ว!", 44 | "Source x Target Mapper is already open.": "ตัวจับคู่ต้นทาง x ปลายทาง เปิดอยู่แล้ว" 45 | } -------------------------------------------------------------------------------- /locales/zh.json: -------------------------------------------------------------------------------- 1 | { 2 | "Source x Target Mapper": "Source x Target Mapper", 3 | "select a source image": "选择一个源图像", 4 | "Preview": "预览", 5 | "select a target image or video": "选择一个目标图像或视频", 6 | "save image output file": "保存图像输出文件", 7 | "save video output file": "保存视频输出文件", 8 | "select a target image": "选择一个目标图像", 9 | "source": "源", 10 | "Select a target": "选择一个目标", 11 | "Select a face": "选择一张脸", 12 | "Keep audio": "保留音频", 13 | "Face Enhancer": "面纹增强器", 14 | "Many faces": "多脸", 15 | "Show FPS": "显示帧率", 16 | "Keep fps": "保持帧率", 17 | "Keep frames": "保持帧数", 18 | "Fix Blueish Cam": "修复偏蓝的摄像头", 19 | "Mouth Mask": "口罩", 20 | "Show Mouth Mask Box": "显示口罩盒", 21 | "Start": "开始", 22 | "Live": "直播", 23 | "Destroy": "结束", 24 | "Map faces": "识别人脸", 25 | "Processing...": "处理中...", 26 | "Processing succeed!": "处理成功!", 27 | "Processing ignored!": "处理被忽略!", 28 | "Failed to start camera": "启动相机失败", 29 | "Please complete pop-up or close it.": "请先完成弹出窗口或者关闭它", 30 | "Getting unique faces": "获取独特面部", 31 | "Please select a source image first": "请先选择一个源图像", 32 | "No faces found in target": "目标图像中没有人脸", 33 | "Add": "添加", 34 | "Clear": "清除", 35 | "Submit": "确认", 36 | "Select source image": "请选取源图像", 37 | "Select target image": "请选取目标图像", 38 | "Please provide mapping!": "请提供映射", 39 | "At least 1 source with target is required!": "至少需要一个来源图像与目标图像相关！", 40 | "At least 1 source with target is required!": "至少需要一个来源图像与目标图像相关！", 41 | "Face could not be detected in last upload!": "最近上传的图像中没有检测到人脸!", 42 | "Select Camera:": "选择摄像头", 43 | "All mappings cleared!": "所有映射均已清除!", 44 | "Mappings successfully submitted!": "成功提交映射！", 45 | "Source x Target Mapper is already open.": "源 x 目标映射器已打开。" 46 | } 47 | -------------------------------------------------------------------------------- /media/Download.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/hacksider/Deep-Live-Cam/ec6d7d2995431fc01cefd17fd4d8a87971169a0e/media/Download.png -------------------------------------------------------------------------------- /media/avgpcperformancedemo.gif: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/hacksider/Deep-Live-Cam/ec6d7d2995431fc01cefd17fd4d8a87971169a0e/media/avgpcperformancedemo.gif -------------------------------------------------------------------------------- /media/deepwarebench.gif: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/hacksider/Deep-Live-Cam/ec6d7d2995431fc01cefd17fd4d8a87971169a0e/media/deepwarebench.gif -------------------------------------------------------------------------------- /media/demo.gif: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/hacksider/Deep-Live-Cam/ec6d7d2995431fc01cefd17fd4d8a87971169a0e/media/demo.gif -------------------------------------------------------------------------------- /media/instruction.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/hacksider/Deep-Live-Cam/ec6d7d2995431fc01cefd17fd4d8a87971169a0e/media/instruction.png -------------------------------------------------------------------------------- /media/live_show.gif: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/hacksider/Deep-Live-Cam/ec6d7d2995431fc01cefd17fd4d8a87971169a0e/media/live_show.gif -------------------------------------------------------------------------------- /media/ludwig.gif: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/hacksider/Deep-Live-Cam/ec6d7d2995431fc01cefd17fd4d8a87971169a0e/media/ludwig.gif -------------------------------------------------------------------------------- /media/meme.gif: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/hacksider/Deep-Live-Cam/ec6d7d2995431fc01cefd17fd4d8a87971169a0e/media/meme.gif -------------------------------------------------------------------------------- /media/movie.gif: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/hacksider/Deep-Live-Cam/ec6d7d2995431fc01cefd17fd4d8a87971169a0e/media/movie.gif -------------------------------------------------------------------------------- /media/streamers.gif: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/hacksider/Deep-Live-Cam/ec6d7d2995431fc01cefd17fd4d8a87971169a0e/media/streamers.gif -------------------------------------------------------------------------------- /models/instructions.txt: -------------------------------------------------------------------------------- 1 | just put the models in this folder - 2 | 3 | https://huggingface.co/hacksider/deep-live-cam/resolve/main/inswapper_128_fp16.onnx?download=true 4 | https://github.com/TencentARC/GFPGAN/releases/download/v1.3.4/GFPGANv1.4.pth 5 | -------------------------------------------------------------------------------- /modules/__init__.py: -------------------------------------------------------------------------------- 1 | import os 2 | import cv2 3 | import numpy as np 4 | 5 | # Utility function to support unicode characters in file paths for reading 6 | def imread_unicode(path, flags=cv2.IMREAD_COLOR): 7 | return cv2.imdecode(np.fromfile(path, dtype=np.uint8), flags) 8 | 9 | # Utility function to support unicode characters in file paths for writing 10 | def imwrite_unicode(path, img, params=None): 11 | root, ext = os.path.splitext(path) 12 | if not ext: 13 | ext = ".png" 14 | result, encoded_img = cv2.imencode(ext, img, params if params else []) 15 | result, encoded_img = cv2.imencode(f".{ext}", img, params if params is not None else []) 16 | encoded_img.tofile(path) 17 | return True 18 | return False -------------------------------------------------------------------------------- /modules/capturer.py: -------------------------------------------------------------------------------- 1 | from typing import Any 2 | import cv2 3 | import modules.globals # Import the globals to check the color correction toggle 4 | 5 | 6 | def get_video_frame(video_path: str, frame_number: int = 0) -> Any: 7 | capture = cv2.VideoCapture(video_path) 8 | 9 | # Set MJPEG format to ensure correct color space handling 10 | capture.set(cv2.CAP_PROP_FOURCC, cv2.VideoWriter_fourcc(*'MJPG')) 11 | 12 | # Only force RGB conversion if color correction is enabled 13 | if modules.globals.color_correction: 14 | capture.set(cv2.CAP_PROP_CONVERT_RGB, 1) 15 | 16 | frame_total = capture.get(cv2.CAP_PROP_FRAME_COUNT) 17 | capture.set(cv2.CAP_PROP_POS_FRAMES, min(frame_total, frame_number - 1)) 18 | has_frame, frame = capture.read() 19 | 20 | if has_frame and modules.globals.color_correction: 21 | # Convert the frame color if necessary 22 | frame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB) 23 | 24 | capture.release() 25 | return frame if has_frame else None 26 | 27 | 28 | def get_video_frame_total(video_path: str) -> int: 29 | capture = cv2.VideoCapture(video_path) 30 | video_frame_total = int(capture.get(cv2.CAP_PROP_FRAME_COUNT)) 31 | capture.release() 32 | return video_frame_total 33 | -------------------------------------------------------------------------------- /modules/cluster_analysis.py: -------------------------------------------------------------------------------- 1 | import numpy as np 2 | from sklearn.cluster import KMeans 3 | from sklearn.metrics import silhouette_score 4 | from typing import Any 5 | 6 | 7 | def find_cluster_centroids(embeddings, max_k=10) -> Any: 8 | inertia = [] 9 | cluster_centroids = [] 10 | K = range(1, max_k+1) 11 | 12 | for k in K: 13 | kmeans = KMeans(n_clusters=k, random_state=0) 14 | kmeans.fit(embeddings) 15 | inertia.append(kmeans.inertia_) 16 | cluster_centroids.append({"k": k, "centroids": kmeans.cluster_centers_}) 17 | 18 | diffs = [inertia[i] - inertia[i+1] for i in range(len(inertia)-1)] 19 | optimal_centroids = cluster_centroids[diffs.index(max(diffs)) + 1]['centroids'] 20 | 21 | return optimal_centroids 22 | 23 | def find_closest_centroid(centroids: list, normed_face_embedding) -> list: 24 | try: 25 | centroids = np.array(centroids) 26 | normed_face_embedding = np.array(normed_face_embedding) 27 | similarities = np.dot(centroids, normed_face_embedding) 28 | closest_centroid_index = np.argmax(similarities) 29 | 30 | return closest_centroid_index, centroids[closest_centroid_index] 31 | except ValueError: 32 | return None -------------------------------------------------------------------------------- /modules/core.py: -------------------------------------------------------------------------------- 1 | import os 2 | import sys 3 | # single thread doubles cuda performance - needs to be set before torch import 4 | if any(arg.startswith('--execution-provider') for arg in sys.argv): 5 | os.environ['OMP_NUM_THREADS'] = '1' 6 | # reduce tensorflow log level 7 | os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2' 8 | import warnings 9 | from typing import List 10 | import platform 11 | import signal 12 | import shutil 13 | import argparse 14 | import torch 15 | import onnxruntime 16 | import tensorflow 17 | 18 | import modules.globals 19 | import modules.metadata 20 | import modules.ui as ui 21 | from modules.processors.frame.core import get_frame_processors_modules 22 | from modules.utilities import has_image_extension, is_image, is_video, detect_fps, create_video, extract_frames, get_temp_frame_paths, restore_audio, create_temp, move_temp, clean_temp, normalize_output_path 23 | 24 | if 'ROCMExecutionProvider' in modules.globals.execution_providers: 25 | del torch 26 | 27 | warnings.filterwarnings('ignore', category=FutureWarning, module='insightface') 28 | warnings.filterwarnings('ignore', category=UserWarning, module='torchvision') 29 | 30 | 31 | def parse_args() -> None: 32 | signal.signal(signal.SIGINT, lambda signal_number, frame: destroy()) 33 | program = argparse.ArgumentParser() 34 | program.add_argument('-s', '--source', help='select an source image', dest='source_path') 35 | program.add_argument('-t', '--target', help='select an target image or video', dest='target_path') 36 | program.add_argument('-o', '--output', help='select output file or directory', dest='output_path') 37 | program.add_argument('--frame-processor', help='pipeline of frame processors', dest='frame_processor', default=['face_swapper'], choices=['face_swapper', 'face_enhancer'], nargs='+') 38 | program.add_argument('--keep-fps', help='keep original fps', dest='keep_fps', action='store_true', default=False) 39 | program.add_argument('--keep-audio', help='keep original audio', dest='keep_audio', action='store_true', default=True) 40 | program.add_argument('--keep-frames', help='keep temporary frames', dest='keep_frames', action='store_true', default=False) 41 | program.add_argument('--many-faces', help='process every face', dest='many_faces', action='store_true', default=False) 42 | program.add_argument('--nsfw-filter', help='filter the NSFW image or video', dest='nsfw_filter', action='store_true', default=False) 43 | program.add_argument('--map-faces', help='map source target faces', dest='map_faces', action='store_true', default=False) 44 | program.add_argument('--mouth-mask', help='mask the mouth region', dest='mouth_mask', action='store_true', default=False) 45 | program.add_argument('--video-encoder', help='adjust output video encoder', dest='video_encoder', default='libx264', choices=['libx264', 'libx265', 'libvpx-vp9']) 46 | program.add_argument('--video-quality', help='adjust output video quality', dest='video_quality', type=int, default=18, choices=range(52), metavar='[0-51]') 47 | program.add_argument('-l', '--lang', help='Ui language', default="en") 48 | program.add_argument('--live-mirror', help='The live camera display as you see it in the front-facing camera frame', dest='live_mirror', action='store_true', default=False) 49 | program.add_argument('--live-resizable', help='The live camera frame is resizable', dest='live_resizable', action='store_true', default=False) 50 | program.add_argument('--max-memory', help='maximum amount of RAM in GB', dest='max_memory', type=int, default=suggest_max_memory()) 51 | program.add_argument('--execution-provider', help='execution provider', dest='execution_provider', default=['cpu'], choices=suggest_execution_providers(), nargs='+') 52 | program.add_argument('--execution-threads', help='number of execution threads', dest='execution_threads', type=int, default=suggest_execution_threads()) 53 | program.add_argument('-v', '--version', action='version', version=f'{modules.metadata.name} {modules.metadata.version}') 54 | 55 | # register deprecated args 56 | program.add_argument('-f', '--face', help=argparse.SUPPRESS, dest='source_path_deprecated') 57 | program.add_argument('--cpu-cores', help=argparse.SUPPRESS, dest='cpu_cores_deprecated', type=int) 58 | program.add_argument('--gpu-vendor', help=argparse.SUPPRESS, dest='gpu_vendor_deprecated') 59 | program.add_argument('--gpu-threads', help=argparse.SUPPRESS, dest='gpu_threads_deprecated', type=int) 60 | 61 | args = program.parse_args() 62 | 63 | modules.globals.source_path = args.source_path 64 | modules.globals.target_path = args.target_path 65 | modules.globals.output_path = normalize_output_path(modules.globals.source_path, modules.globals.target_path, args.output_path) 66 | modules.globals.frame_processors = args.frame_processor 67 | modules.globals.headless = args.source_path or args.target_path or args.output_path 68 | modules.globals.keep_fps = args.keep_fps 69 | modules.globals.keep_audio = args.keep_audio 70 | modules.globals.keep_frames = args.keep_frames 71 | modules.globals.many_faces = args.many_faces 72 | modules.globals.mouth_mask = args.mouth_mask 73 | modules.globals.nsfw_filter = args.nsfw_filter 74 | modules.globals.map_faces = args.map_faces 75 | modules.globals.video_encoder = args.video_encoder 76 | modules.globals.video_quality = args.video_quality 77 | modules.globals.live_mirror = args.live_mirror 78 | modules.globals.live_resizable = args.live_resizable 79 | modules.globals.max_memory = args.max_memory 80 | modules.globals.execution_providers = decode_execution_providers(args.execution_provider) 81 | modules.globals.execution_threads = args.execution_threads 82 | modules.globals.lang = args.lang 83 | 84 | #for ENHANCER tumbler: 85 | if 'face_enhancer' in args.frame_processor: 86 | modules.globals.fp_ui['face_enhancer'] = True 87 | else: 88 | modules.globals.fp_ui['face_enhancer'] = False 89 | 90 | # translate deprecated args 91 | if args.source_path_deprecated: 92 | print('\033[33mArgument -f and --face are deprecated. Use -s and --source instead.\033[0m') 93 | modules.globals.source_path = args.source_path_deprecated 94 | modules.globals.output_path = normalize_output_path(args.source_path_deprecated, modules.globals.target_path, args.output_path) 95 | if args.cpu_cores_deprecated: 96 | print('\033[33mArgument --cpu-cores is deprecated. Use --execution-threads instead.\033[0m') 97 | modules.globals.execution_threads = args.cpu_cores_deprecated 98 | if args.gpu_vendor_deprecated == 'apple': 99 | print('\033[33mArgument --gpu-vendor apple is deprecated. Use --execution-provider coreml instead.\033[0m') 100 | modules.globals.execution_providers = decode_execution_providers(['coreml']) 101 | if args.gpu_vendor_deprecated == 'nvidia': 102 | print('\033[33mArgument --gpu-vendor nvidia is deprecated. Use --execution-provider cuda instead.\033[0m') 103 | modules.globals.execution_providers = decode_execution_providers(['cuda']) 104 | if args.gpu_vendor_deprecated == 'amd': 105 | print('\033[33mArgument --gpu-vendor amd is deprecated. Use --execution-provider cuda instead.\033[0m') 106 | modules.globals.execution_providers = decode_execution_providers(['rocm']) 107 | if args.gpu_threads_deprecated: 108 | print('\033[33mArgument --gpu-threads is deprecated. Use --execution-threads instead.\033[0m') 109 | modules.globals.execution_threads = args.gpu_threads_deprecated 110 | 111 | 112 | def encode_execution_providers(execution_providers: List[str]) -> List[str]: 113 | return [execution_provider.replace('ExecutionProvider', '').lower() for execution_provider in execution_providers] 114 | 115 | 116 | def decode_execution_providers(execution_providers: List[str]) -> List[str]: 117 | return [provider for provider, encoded_execution_provider in zip(onnxruntime.get_available_providers(), encode_execution_providers(onnxruntime.get_available_providers())) 118 | if any(execution_provider in encoded_execution_provider for execution_provider in execution_providers)] 119 | 120 | 121 | def suggest_max_memory() -> int: 122 | if platform.system().lower() == 'darwin': 123 | return 4 124 | return 16 125 | 126 | 127 | def suggest_execution_providers() -> List[str]: 128 | return encode_execution_providers(onnxruntime.get_available_providers()) 129 | 130 | 131 | def suggest_execution_threads() -> int: 132 | if 'DmlExecutionProvider' in modules.globals.execution_providers: 133 | return 1 134 | if 'ROCMExecutionProvider' in modules.globals.execution_providers: 135 | return 1 136 | return 8 137 | 138 | 139 | def limit_resources() -> None: 140 | # prevent tensorflow memory leak 141 | gpus = tensorflow.config.experimental.list_physical_devices('GPU') 142 | for gpu in gpus: 143 | tensorflow.config.experimental.set_memory_growth(gpu, True) 144 | # limit memory usage 145 | if modules.globals.max_memory: 146 | memory = modules.globals.max_memory * 1024 ** 3 147 | if platform.system().lower() == 'darwin': 148 | memory = modules.globals.max_memory * 1024 ** 6 149 | if platform.system().lower() == 'windows': 150 | import ctypes 151 | kernel32 = ctypes.windll.kernel32 152 | kernel32.SetProcessWorkingSetSize(-1, ctypes.c_size_t(memory), ctypes.c_size_t(memory)) 153 | else: 154 | import resource 155 | resource.setrlimit(resource.RLIMIT_DATA, (memory, memory)) 156 | 157 | 158 | def release_resources() -> None: 159 | if 'CUDAExecutionProvider' in modules.globals.execution_providers: 160 | torch.cuda.empty_cache() 161 | 162 | 163 | def pre_check() -> bool: 164 | if sys.version_info < (3, 9): 165 | update_status('Python version is not supported - please upgrade to 3.9 or higher.') 166 | return False 167 | if not shutil.which('ffmpeg'): 168 | update_status('ffmpeg is not installed.') 169 | return False 170 | return True 171 | 172 | 173 | def update_status(message: str, scope: str = 'DLC.CORE') -> None: 174 | print(f'[{scope}] {message}') 175 | if not modules.globals.headless: 176 | ui.update_status(message) 177 | 178 | def start() -> None: 179 | for frame_processor in get_frame_processors_modules(modules.globals.frame_processors): 180 | if not frame_processor.pre_start(): 181 | return 182 | update_status('Processing...') 183 | # process image to image 184 | if has_image_extension(modules.globals.target_path): 185 | if modules.globals.nsfw_filter and ui.check_and_ignore_nsfw(modules.globals.target_path, destroy): 186 | return 187 | try: 188 | shutil.copy2(modules.globals.target_path, modules.globals.output_path) 189 | except Exception as e: 190 | print("Error copying file:", str(e)) 191 | for frame_processor in get_frame_processors_modules(modules.globals.frame_processors): 192 | update_status('Progressing...', frame_processor.NAME) 193 | frame_processor.process_image(modules.globals.source_path, modules.globals.output_path, modules.globals.output_path) 194 | release_resources() 195 | if is_image(modules.globals.target_path): 196 | update_status('Processing to image succeed!') 197 | else: 198 | update_status('Processing to image failed!') 199 | return 200 | # process image to videos 201 | if modules.globals.nsfw_filter and ui.check_and_ignore_nsfw(modules.globals.target_path, destroy): 202 | return 203 | 204 | if not modules.globals.map_faces: 205 | update_status('Creating temp resources...') 206 | create_temp(modules.globals.target_path) 207 | update_status('Extracting frames...') 208 | extract_frames(modules.globals.target_path) 209 | 210 | temp_frame_paths = get_temp_frame_paths(modules.globals.target_path) 211 | for frame_processor in get_frame_processors_modules(modules.globals.frame_processors): 212 | update_status('Progressing...', frame_processor.NAME) 213 | frame_processor.process_video(modules.globals.source_path, temp_frame_paths) 214 | release_resources() 215 | # handles fps 216 | if modules.globals.keep_fps: 217 | update_status('Detecting fps...') 218 | fps = detect_fps(modules.globals.target_path) 219 | update_status(f'Creating video with {fps} fps...') 220 | create_video(modules.globals.target_path, fps) 221 | else: 222 | update_status('Creating video with 30.0 fps...') 223 | create_video(modules.globals.target_path) 224 | # handle audio 225 | if modules.globals.keep_audio: 226 | if modules.globals.keep_fps: 227 | update_status('Restoring audio...') 228 | else: 229 | update_status('Restoring audio might cause issues as fps are not kept...') 230 | restore_audio(modules.globals.target_path, modules.globals.output_path) 231 | else: 232 | move_temp(modules.globals.target_path, modules.globals.output_path) 233 | # clean and validate 234 | clean_temp(modules.globals.target_path) 235 | if is_video(modules.globals.target_path): 236 | update_status('Processing to video succeed!') 237 | else: 238 | update_status('Processing to video failed!') 239 | 240 | 241 | def destroy(to_quit=True) -> None: 242 | if modules.globals.target_path: 243 | clean_temp(modules.globals.target_path) 244 | if to_quit: quit() 245 | 246 | 247 | def run() -> None: 248 | parse_args() 249 | if not pre_check(): 250 | return 251 | for frame_processor in get_frame_processors_modules(modules.globals.frame_processors): 252 | if not frame_processor.pre_check(): 253 | return 254 | limit_resources() 255 | if modules.globals.headless: 256 | start() 257 | else: 258 | window = ui.init(start, destroy, modules.globals.lang) 259 | window.mainloop() 260 | -------------------------------------------------------------------------------- /modules/face_analyser.py: -------------------------------------------------------------------------------- 1 | import os 2 | import shutil 3 | from typing import Any 4 | import insightface 5 | 6 | import cv2 7 | import numpy as np 8 | import modules.globals 9 | from tqdm import tqdm 10 | from modules.typing import Frame 11 | from modules.cluster_analysis import find_cluster_centroids, find_closest_centroid 12 | from modules.utilities import get_temp_directory_path, create_temp, extract_frames, clean_temp, get_temp_frame_paths 13 | from pathlib import Path 14 | 15 | FACE_ANALYSER = None 16 | 17 | 18 | def get_face_analyser() -> Any: 19 | global FACE_ANALYSER 20 | 21 | if FACE_ANALYSER is None: 22 | FACE_ANALYSER = insightface.app.FaceAnalysis(name='buffalo_l', providers=modules.globals.execution_providers) 23 | FACE_ANALYSER.prepare(ctx_id=0, det_size=(640, 640)) 24 | return FACE_ANALYSER 25 | 26 | 27 | def get_one_face(frame: Frame) -> Any: 28 | face = get_face_analyser().get(frame) 29 | try: 30 | return min(face, key=lambda x: x.bbox[0]) 31 | except ValueError: 32 | return None 33 | 34 | 35 | def get_many_faces(frame: Frame) -> Any: 36 | try: 37 | return get_face_analyser().get(frame) 38 | except IndexError: 39 | return None 40 | 41 | def has_valid_map() -> bool: 42 | for map in modules.globals.source_target_map: 43 | if "source" in map and "target" in map: 44 | return True 45 | return False 46 | 47 | def default_source_face() -> Any: 48 | for map in modules.globals.source_target_map: 49 | if "source" in map: 50 | return map['source']['face'] 51 | return None 52 | 53 | def simplify_maps() -> Any: 54 | centroids = [] 55 | faces = [] 56 | for map in modules.globals.source_target_map: 57 | if "source" in map and "target" in map: 58 | centroids.append(map['target']['face'].normed_embedding) 59 | faces.append(map['source']['face']) 60 | 61 | modules.globals.simple_map = {'source_faces': faces, 'target_embeddings': centroids} 62 | return None 63 | 64 | def add_blank_map() -> Any: 65 | try: 66 | max_id = -1 67 | if len(modules.globals.source_target_map) > 0: 68 | max_id = max(modules.globals.source_target_map, key=lambda x: x['id'])['id'] 69 | 70 | modules.globals.source_target_map.append({ 71 | 'id' : max_id + 1 72 | }) 73 | except ValueError: 74 | return None 75 | 76 | def get_unique_faces_from_target_image() -> Any: 77 | try: 78 | modules.globals.source_target_map = [] 79 | target_frame = cv2.imread(modules.globals.target_path) 80 | many_faces = get_many_faces(target_frame) 81 | i = 0 82 | 83 | for face in many_faces: 84 | x_min, y_min, x_max, y_max = face['bbox'] 85 | modules.globals.source_target_map.append({ 86 | 'id' : i, 87 | 'target' : { 88 | 'cv2' : target_frame[int(y_min):int(y_max), int(x_min):int(x_max)], 89 | 'face' : face 90 | } 91 | }) 92 | i = i + 1 93 | except ValueError: 94 | return None 95 | 96 | 97 | def get_unique_faces_from_target_video() -> Any: 98 | try: 99 | modules.globals.source_target_map = [] 100 | frame_face_embeddings = [] 101 | face_embeddings = [] 102 | 103 | print('Creating temp resources...') 104 | clean_temp(modules.globals.target_path) 105 | create_temp(modules.globals.target_path) 106 | print('Extracting frames...') 107 | extract_frames(modules.globals.target_path) 108 | 109 | temp_frame_paths = get_temp_frame_paths(modules.globals.target_path) 110 | 111 | i = 0 112 | for temp_frame_path in tqdm(temp_frame_paths, desc="Extracting face embeddings from frames"): 113 | temp_frame = cv2.imread(temp_frame_path) 114 | many_faces = get_many_faces(temp_frame) 115 | 116 | for face in many_faces: 117 | face_embeddings.append(face.normed_embedding) 118 | 119 | frame_face_embeddings.append({'frame': i, 'faces': many_faces, 'location': temp_frame_path}) 120 | i += 1 121 | 122 | centroids = find_cluster_centroids(face_embeddings) 123 | 124 | for frame in frame_face_embeddings: 125 | for face in frame['faces']: 126 | closest_centroid_index, _ = find_closest_centroid(centroids, face.normed_embedding) 127 | face['target_centroid'] = closest_centroid_index 128 | 129 | for i in range(len(centroids)): 130 | modules.globals.source_target_map.append({ 131 | 'id' : i 132 | }) 133 | 134 | temp = [] 135 | for frame in tqdm(frame_face_embeddings, desc=f"Mapping frame embeddings to centroids-{i}"): 136 | temp.append({'frame': frame['frame'], 'faces': [face for face in frame['faces'] if face['target_centroid'] == i], 'location': frame['location']}) 137 | 138 | modules.globals.source_target_map[i]['target_faces_in_frame'] = temp 139 | 140 | # dump_faces(centroids, frame_face_embeddings) 141 | default_target_face() 142 | except ValueError: 143 | return None 144 | 145 | 146 | def default_target_face(): 147 | for map in modules.globals.source_target_map: 148 | best_face = None 149 | best_frame = None 150 | for frame in map['target_faces_in_frame']: 151 | if len(frame['faces']) > 0: 152 | best_face = frame['faces'][0] 153 | best_frame = frame 154 | break 155 | 156 | for frame in map['target_faces_in_frame']: 157 | for face in frame['faces']: 158 | if face['det_score'] > best_face['det_score']: 159 | best_face = face 160 | best_frame = frame 161 | 162 | x_min, y_min, x_max, y_max = best_face['bbox'] 163 | 164 | target_frame = cv2.imread(best_frame['location']) 165 | map['target'] = { 166 | 'cv2' : target_frame[int(y_min):int(y_max), int(x_min):int(x_max)], 167 | 'face' : best_face 168 | } 169 | 170 | 171 | def dump_faces(centroids: Any, frame_face_embeddings: list): 172 | temp_directory_path = get_temp_directory_path(modules.globals.target_path) 173 | 174 | for i in range(len(centroids)): 175 | if os.path.exists(temp_directory_path + f"/{i}") and os.path.isdir(temp_directory_path + f"/{i}"): 176 | shutil.rmtree(temp_directory_path + f"/{i}") 177 | Path(temp_directory_path + f"/{i}").mkdir(parents=True, exist_ok=True) 178 | 179 | for frame in tqdm(frame_face_embeddings, desc=f"Copying faces to temp/./{i}"): 180 | temp_frame = cv2.imread(frame['location']) 181 | 182 | j = 0 183 | for face in frame['faces']: 184 | if face['target_centroid'] == i: 185 | x_min, y_min, x_max, y_max = face['bbox'] 186 | 187 | if temp_frame[int(y_min):int(y_max), int(x_min):int(x_max)].size > 0: 188 | cv2.imwrite(temp_directory_path + f"/{i}/{frame['frame']}_{j}.png", temp_frame[int(y_min):int(y_max), int(x_min):int(x_max)]) 189 | j += 1 -------------------------------------------------------------------------------- /modules/gettext.py: -------------------------------------------------------------------------------- 1 | import json 2 | from pathlib import Path 3 | 4 | class LanguageManager: 5 | def __init__(self, default_language="en"): 6 | self.current_language = default_language 7 | self.translations = {} 8 | self.load_language(default_language) 9 | 10 | def load_language(self, language_code) -> bool: 11 | """load language file""" 12 | if language_code == "en": 13 | return True 14 | try: 15 | file_path = Path(__file__).parent.parent / f"locales/{language_code}.json" 16 | with open(file_path, "r", encoding="utf-8") as file: 17 | self.translations = json.load(file) 18 | self.current_language = language_code 19 | return True 20 | except FileNotFoundError: 21 | print(f"Language file not found: {language_code}") 22 | return False 23 | 24 | def _(self, key, default=None) -> str: 25 | """get translate text""" 26 | return self.translations.get(key, default if default else key) -------------------------------------------------------------------------------- /modules/globals.py: -------------------------------------------------------------------------------- 1 | import os 2 | from typing import List, Dict, Any 3 | 4 | ROOT_DIR = os.path.dirname(os.path.abspath(__file__)) 5 | WORKFLOW_DIR = os.path.join(ROOT_DIR, "workflow") 6 | 7 | file_types = [ 8 | ("Image", ("*.png", "*.jpg", "*.jpeg", "*.gif", "*.bmp")), 9 | ("Video", ("*.mp4", "*.mkv")), 10 | ] 11 | 12 | source_target_map = [] 13 | simple_map = {} 14 | 15 | source_path = None 16 | target_path = None 17 | output_path = None 18 | frame_processors: List[str] = [] 19 | keep_fps = True 20 | keep_audio = True 21 | keep_frames = False 22 | many_faces = False 23 | map_faces = False 24 | color_correction = False # New global variable for color correction toggle 25 | nsfw_filter = False 26 | video_encoder = None 27 | video_quality = None 28 | live_mirror = False 29 | live_resizable = True 30 | max_memory = None 31 | execution_providers: List[str] = [] 32 | execution_threads = None 33 | headless = None 34 | log_level = "error" 35 | fp_ui: Dict[str, bool] = {"face_enhancer": False} 36 | camera_input_combobox = None 37 | webcam_preview_running = False 38 | show_fps = False 39 | mouth_mask = False 40 | show_mouth_mask_box = False 41 | mask_feather_ratio = 8 42 | mask_down_size = 0.50 43 | mask_size = 1 44 | -------------------------------------------------------------------------------- /modules/metadata.py: -------------------------------------------------------------------------------- 1 | name = 'Deep-Live-Cam' 2 | version = '1.8' 3 | edition = 'GitHub Edition' 4 | -------------------------------------------------------------------------------- /modules/predicter.py: -------------------------------------------------------------------------------- 1 | import numpy 2 | import opennsfw2 3 | from PIL import Image 4 | import cv2 # Add OpenCV import 5 | import modules.globals # Import globals to access the color correction toggle 6 | 7 | from modules.typing import Frame 8 | 9 | MAX_PROBABILITY = 0.85 10 | 11 | # Preload the model once for efficiency 12 | model = None 13 | 14 | def predict_frame(target_frame: Frame) -> bool: 15 | # Convert the frame to RGB before processing if color correction is enabled 16 | if modules.globals.color_correction: 17 | target_frame = cv2.cvtColor(target_frame, cv2.COLOR_BGR2RGB) 18 | 19 | image = Image.fromarray(target_frame) 20 | image = opennsfw2.preprocess_image(image, opennsfw2.Preprocessing.YAHOO) 21 | global model 22 | if model is None: 23 | model = opennsfw2.make_open_nsfw_model() 24 | 25 | views = numpy.expand_dims(image, axis=0) 26 | _, probability = model.predict(views)[0] 27 | return probability > MAX_PROBABILITY 28 | 29 | 30 | def predict_image(target_path: str) -> bool: 31 | return opennsfw2.predict_image(target_path) > MAX_PROBABILITY 32 | 33 | 34 | def predict_video(target_path: str) -> bool: 35 | _, probabilities = opennsfw2.predict_video_frames(video_path=target_path, frame_interval=100) 36 | return any(probability > MAX_PROBABILITY for probability in probabilities) 37 | -------------------------------------------------------------------------------- /modules/processors/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/hacksider/Deep-Live-Cam/ec6d7d2995431fc01cefd17fd4d8a87971169a0e/modules/processors/__init__.py -------------------------------------------------------------------------------- /modules/processors/frame/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/hacksider/Deep-Live-Cam/ec6d7d2995431fc01cefd17fd4d8a87971169a0e/modules/processors/frame/__init__.py -------------------------------------------------------------------------------- /modules/processors/frame/core.py: -------------------------------------------------------------------------------- 1 | import sys 2 | import importlib 3 | from concurrent.futures import ThreadPoolExecutor 4 | from types import ModuleType 5 | from typing import Any, List, Callable 6 | from tqdm import tqdm 7 | 8 | import modules 9 | import modules.globals 10 | 11 | FRAME_PROCESSORS_MODULES: List[ModuleType] = [] 12 | FRAME_PROCESSORS_INTERFACE = [ 13 | 'pre_check', 14 | 'pre_start', 15 | 'process_frame', 16 | 'process_image', 17 | 'process_video' 18 | ] 19 | 20 | 21 | def load_frame_processor_module(frame_processor: str) -> Any: 22 | try: 23 | frame_processor_module = importlib.import_module(f'modules.processors.frame.{frame_processor}') 24 | for method_name in FRAME_PROCESSORS_INTERFACE: 25 | if not hasattr(frame_processor_module, method_name): 26 | sys.exit() 27 | except ImportError: 28 | print(f"Frame processor {frame_processor} not found") 29 | sys.exit() 30 | return frame_processor_module 31 | 32 | 33 | def get_frame_processors_modules(frame_processors: List[str]) -> List[ModuleType]: 34 | global FRAME_PROCESSORS_MODULES 35 | 36 | if not FRAME_PROCESSORS_MODULES: 37 | for frame_processor in frame_processors: 38 | frame_processor_module = load_frame_processor_module(frame_processor) 39 | FRAME_PROCESSORS_MODULES.append(frame_processor_module) 40 | set_frame_processors_modules_from_ui(frame_processors) 41 | return FRAME_PROCESSORS_MODULES 42 | 43 | def set_frame_processors_modules_from_ui(frame_processors: List[str]) -> None: 44 | global FRAME_PROCESSORS_MODULES 45 | current_processor_names = [proc.__name__.split('.')[-1] for proc in FRAME_PROCESSORS_MODULES] 46 | 47 | for frame_processor, state in modules.globals.fp_ui.items(): 48 | if state == True and frame_processor not in current_processor_names: 49 | try: 50 | frame_processor_module = load_frame_processor_module(frame_processor) 51 | FRAME_PROCESSORS_MODULES.append(frame_processor_module) 52 | if frame_processor not in modules.globals.frame_processors: 53 | modules.globals.frame_processors.append(frame_processor) 54 | except SystemExit: 55 | print(f"Warning: Failed to load frame processor {frame_processor} requested by UI state.") 56 | except Exception as e: 57 | print(f"Warning: Error loading frame processor {frame_processor} requested by UI state: {e}") 58 | 59 | elif state == False and frame_processor in current_processor_names: 60 | try: 61 | module_to_remove = next((mod for mod in FRAME_PROCESSORS_MODULES if mod.__name__.endswith(f'.{frame_processor}')), None) 62 | if module_to_remove: 63 | FRAME_PROCESSORS_MODULES.remove(module_to_remove) 64 | if frame_processor in modules.globals.frame_processors: 65 | modules.globals.frame_processors.remove(frame_processor) 66 | except Exception as e: 67 | print(f"Warning: Error removing frame processor {frame_processor}: {e}") 68 | 69 | def multi_process_frame(source_path: str, temp_frame_paths: List[str], process_frames: Callable[[str, List[str], Any], None], progress: Any = None) -> None: 70 | with ThreadPoolExecutor(max_workers=modules.globals.execution_threads) as executor: 71 | futures = [] 72 | for path in temp_frame_paths: 73 | future = executor.submit(process_frames, source_path, [path], progress) 74 | futures.append(future) 75 | for future in futures: 76 | future.result() 77 | 78 | 79 | def process_video(source_path: str, frame_paths: list[str], process_frames: Callable[[str, List[str], Any], None]) -> None: 80 | progress_bar_format = '{l_bar}{bar}| {n_fmt}/{total_fmt} [{elapsed}<{remaining}, {rate_fmt}{postfix}]' 81 | total = len(frame_paths) 82 | with tqdm(total=total, desc='Processing', unit='frame', dynamic_ncols=True, bar_format=progress_bar_format) as progress: 83 | progress.set_postfix({'execution_providers': modules.globals.execution_providers, 'execution_threads': modules.globals.execution_threads, 'max_memory': modules.globals.max_memory}) 84 | multi_process_frame(source_path, frame_paths, process_frames, progress) 85 | -------------------------------------------------------------------------------- /modules/processors/frame/face_enhancer.py: -------------------------------------------------------------------------------- 1 | from typing import Any, List 2 | import cv2 3 | import threading 4 | import gfpgan 5 | import os 6 | 7 | import modules.globals 8 | import modules.processors.frame.core 9 | from modules.core import update_status 10 | from modules.face_analyser import get_one_face 11 | from modules.typing import Frame, Face 12 | import platform 13 | import torch 14 | from modules.utilities import ( 15 | conditional_download, 16 | is_image, 17 | is_video, 18 | ) 19 | 20 | FACE_ENHANCER = None 21 | THREAD_SEMAPHORE = threading.Semaphore() 22 | THREAD_LOCK = threading.Lock() 23 | NAME = "DLC.FACE-ENHANCER" 24 | 25 | abs_dir = os.path.dirname(os.path.abspath(__file__)) 26 | models_dir = os.path.join( 27 | os.path.dirname(os.path.dirname(os.path.dirname(abs_dir))), "models" 28 | ) 29 | 30 | 31 | def pre_check() -> bool: 32 | download_directory_path = models_dir 33 | conditional_download( 34 | download_directory_path, 35 | [ 36 | "https://github.com/TencentARC/GFPGAN/releases/download/v1.3.4/GFPGANv1.4.pth" 37 | ], 38 | ) 39 | return True 40 | 41 | 42 | def pre_start() -> bool: 43 | if not is_image(modules.globals.target_path) and not is_video( 44 | modules.globals.target_path 45 | ): 46 | update_status("Select an image or video for target path.", NAME) 47 | return False 48 | return True 49 | 50 | 51 | TENSORRT_AVAILABLE = False 52 | try: 53 | import torch_tensorrt 54 | TENSORRT_AVAILABLE = True 55 | except ImportError as im: 56 | print(f"TensorRT is not available: {im}") 57 | pass 58 | except Exception as e: 59 | print(f"TensorRT is not available: {e}") 60 | pass 61 | 62 | def get_face_enhancer() -> Any: 63 | global FACE_ENHANCER 64 | 65 | with THREAD_LOCK: 66 | if FACE_ENHANCER is None: 67 | model_path = os.path.join(models_dir, "GFPGANv1.4.pth") 68 | 69 | selected_device = None 70 | device_priority = [] 71 | 72 | if TENSORRT_AVAILABLE and torch.cuda.is_available(): 73 | selected_device = torch.device("cuda") 74 | device_priority.append("TensorRT+CUDA") 75 | elif torch.cuda.is_available(): 76 | selected_device = torch.device("cuda") 77 | device_priority.append("CUDA") 78 | elif torch.backends.mps.is_available() and platform.system() == "Darwin": 79 | selected_device = torch.device("mps") 80 | device_priority.append("MPS") 81 | elif not torch.cuda.is_available(): 82 | selected_device = torch.device("cpu") 83 | device_priority.append("CPU") 84 | 85 | FACE_ENHANCER = gfpgan.GFPGANer(model_path=model_path, upscale=1, device=selected_device) 86 | 87 | # for debug: 88 | print(f"Selected device: {selected_device} and device priority: {device_priority}") 89 | return FACE_ENHANCER 90 | 91 | 92 | def enhance_face(temp_frame: Frame) -> Frame: 93 | with THREAD_SEMAPHORE: 94 | _, _, temp_frame = get_face_enhancer().enhance(temp_frame, paste_back=True) 95 | return temp_frame 96 | 97 | 98 | def process_frame(source_face: Face, temp_frame: Frame) -> Frame: 99 | target_face = get_one_face(temp_frame) 100 | if target_face: 101 | temp_frame = enhance_face(temp_frame) 102 | return temp_frame 103 | 104 | 105 | def process_frames( 106 | source_path: str, temp_frame_paths: List[str], progress: Any = None 107 | ) -> None: 108 | for temp_frame_path in temp_frame_paths: 109 | temp_frame = cv2.imread(temp_frame_path) 110 | result = process_frame(None, temp_frame) 111 | cv2.imwrite(temp_frame_path, result) 112 | if progress: 113 | progress.update(1) 114 | 115 | 116 | def process_image(source_path: str, target_path: str, output_path: str) -> None: 117 | target_frame = cv2.imread(target_path) 118 | result = process_frame(None, target_frame) 119 | cv2.imwrite(output_path, result) 120 | 121 | 122 | def process_video(source_path: str, temp_frame_paths: List[str]) -> None: 123 | modules.processors.frame.core.process_video(None, temp_frame_paths, process_frames) 124 | 125 | 126 | def process_frame_v2(temp_frame: Frame) -> Frame: 127 | target_face = get_one_face(temp_frame) 128 | if target_face: 129 | temp_frame = enhance_face(temp_frame) 130 | return temp_frame 131 | -------------------------------------------------------------------------------- /modules/processors/frame/face_swapper.py: -------------------------------------------------------------------------------- 1 | from typing import Any, List 2 | import cv2 3 | import insightface 4 | import threading 5 | import numpy as np 6 | import modules.globals 7 | import logging 8 | import modules.processors.frame.core 9 | from modules.core import update_status 10 | from modules.face_analyser import get_one_face, get_many_faces, default_source_face 11 | from modules.typing import Face, Frame 12 | from modules.utilities import ( 13 | conditional_download, 14 | is_image, 15 | is_video, 16 | ) 17 | from modules.cluster_analysis import find_closest_centroid 18 | import os 19 | 20 | FACE_SWAPPER = None 21 | THREAD_LOCK = threading.Lock() 22 | NAME = "DLC.FACE-SWAPPER" 23 | 24 | abs_dir = os.path.dirname(os.path.abspath(__file__)) 25 | models_dir = os.path.join( 26 | os.path.dirname(os.path.dirname(os.path.dirname(abs_dir))), "models" 27 | ) 28 | 29 | 30 | def pre_check() -> bool: 31 | download_directory_path = abs_dir 32 | conditional_download( 33 | download_directory_path, 34 | [ 35 | "https://huggingface.co/hacksider/deep-live-cam/blob/main/inswapper_128_fp16.onnx" 36 | ], 37 | ) 38 | return True 39 | 40 | 41 | def pre_start() -> bool: 42 | if not modules.globals.map_faces and not is_image(modules.globals.source_path): 43 | update_status("Select an image for source path.", NAME) 44 | return False 45 | elif not modules.globals.map_faces and not get_one_face( 46 | cv2.imread(modules.globals.source_path) 47 | ): 48 | update_status("No face in source path detected.", NAME) 49 | return False 50 | if not is_image(modules.globals.target_path) and not is_video( 51 | modules.globals.target_path 52 | ): 53 | update_status("Select an image or video for target path.", NAME) 54 | return False 55 | return True 56 | 57 | 58 | def get_face_swapper() -> Any: 59 | global FACE_SWAPPER 60 | 61 | with THREAD_LOCK: 62 | if FACE_SWAPPER is None: 63 | model_path = os.path.join(models_dir, "inswapper_128_fp16.onnx") 64 | FACE_SWAPPER = insightface.model_zoo.get_model( 65 | model_path, providers=modules.globals.execution_providers 66 | ) 67 | return FACE_SWAPPER 68 | 69 | 70 | def swap_face(source_face: Face, target_face: Face, temp_frame: Frame) -> Frame: 71 | face_swapper = get_face_swapper() 72 | 73 | # Apply the face swap 74 | swapped_frame = face_swapper.get( 75 | temp_frame, target_face, source_face, paste_back=True 76 | ) 77 | 78 | if modules.globals.mouth_mask: 79 | # Create a mask for the target face 80 | face_mask = create_face_mask(target_face, temp_frame) 81 | 82 | # Create the mouth mask 83 | mouth_mask, mouth_cutout, mouth_box, lower_lip_polygon = ( 84 | create_lower_mouth_mask(target_face, temp_frame) 85 | ) 86 | 87 | # Apply the mouth area 88 | swapped_frame = apply_mouth_area( 89 | swapped_frame, mouth_cutout, mouth_box, face_mask, lower_lip_polygon 90 | ) 91 | 92 | if modules.globals.show_mouth_mask_box: 93 | mouth_mask_data = (mouth_mask, mouth_cutout, mouth_box, lower_lip_polygon) 94 | swapped_frame = draw_mouth_mask_visualization( 95 | swapped_frame, target_face, mouth_mask_data 96 | ) 97 | 98 | return swapped_frame 99 | 100 | 101 | def process_frame(source_face: Face, temp_frame: Frame) -> Frame: 102 | if modules.globals.color_correction: 103 | temp_frame = cv2.cvtColor(temp_frame, cv2.COLOR_BGR2RGB) 104 | 105 | if modules.globals.many_faces: 106 | many_faces = get_many_faces(temp_frame) 107 | if many_faces: 108 | for target_face in many_faces: 109 | if source_face and target_face: 110 | temp_frame = swap_face(source_face, target_face, temp_frame) 111 | else: 112 | print("Face detection failed for target/source.") 113 | else: 114 | target_face = get_one_face(temp_frame) 115 | if target_face and source_face: 116 | temp_frame = swap_face(source_face, target_face, temp_frame) 117 | else: 118 | logging.error("Face detection failed for target or source.") 119 | return temp_frame 120 | 121 | 122 | 123 | def process_frame_v2(temp_frame: Frame, temp_frame_path: str = "") -> Frame: 124 | if is_image(modules.globals.target_path): 125 | if modules.globals.many_faces: 126 | source_face = default_source_face() 127 | for map in modules.globals.source_target_map: 128 | target_face = map["target"]["face"] 129 | temp_frame = swap_face(source_face, target_face, temp_frame) 130 | 131 | elif not modules.globals.many_faces: 132 | for map in modules.globals.source_target_map: 133 | if "source" in map: 134 | source_face = map["source"]["face"] 135 | target_face = map["target"]["face"] 136 | temp_frame = swap_face(source_face, target_face, temp_frame) 137 | 138 | elif is_video(modules.globals.target_path): 139 | if modules.globals.many_faces: 140 | source_face = default_source_face() 141 | for map in modules.globals.source_target_map: 142 | target_frame = [ 143 | f 144 | for f in map["target_faces_in_frame"] 145 | if f["location"] == temp_frame_path 146 | ] 147 | 148 | for frame in target_frame: 149 | for target_face in frame["faces"]: 150 | temp_frame = swap_face(source_face, target_face, temp_frame) 151 | 152 | elif not modules.globals.many_faces: 153 | for map in modules.globals.source_target_map: 154 | if "source" in map: 155 | target_frame = [ 156 | f 157 | for f in map["target_faces_in_frame"] 158 | if f["location"] == temp_frame_path 159 | ] 160 | source_face = map["source"]["face"] 161 | 162 | for frame in target_frame: 163 | for target_face in frame["faces"]: 164 | temp_frame = swap_face(source_face, target_face, temp_frame) 165 | 166 | else: 167 | detected_faces = get_many_faces(temp_frame) 168 | if modules.globals.many_faces: 169 | if detected_faces: 170 | source_face = default_source_face() 171 | for target_face in detected_faces: 172 | temp_frame = swap_face(source_face, target_face, temp_frame) 173 | 174 | elif not modules.globals.many_faces: 175 | if detected_faces: 176 | if len(detected_faces) <= len( 177 | modules.globals.simple_map["target_embeddings"] 178 | ): 179 | for detected_face in detected_faces: 180 | closest_centroid_index, _ = find_closest_centroid( 181 | modules.globals.simple_map["target_embeddings"], 182 | detected_face.normed_embedding, 183 | ) 184 | 185 | temp_frame = swap_face( 186 | modules.globals.simple_map["source_faces"][ 187 | closest_centroid_index 188 | ], 189 | detected_face, 190 | temp_frame, 191 | ) 192 | else: 193 | detected_faces_centroids = [] 194 | for face in detected_faces: 195 | detected_faces_centroids.append(face.normed_embedding) 196 | i = 0 197 | for target_embedding in modules.globals.simple_map[ 198 | "target_embeddings" 199 | ]: 200 | closest_centroid_index, _ = find_closest_centroid( 201 | detected_faces_centroids, target_embedding 202 | ) 203 | 204 | temp_frame = swap_face( 205 | modules.globals.simple_map["source_faces"][i], 206 | detected_faces[closest_centroid_index], 207 | temp_frame, 208 | ) 209 | i += 1 210 | return temp_frame 211 | 212 | 213 | def process_frames( 214 | source_path: str, temp_frame_paths: List[str], progress: Any = None 215 | ) -> None: 216 | if not modules.globals.map_faces: 217 | source_face = get_one_face(cv2.imread(source_path)) 218 | for temp_frame_path in temp_frame_paths: 219 | temp_frame = cv2.imread(temp_frame_path) 220 | try: 221 | result = process_frame(source_face, temp_frame) 222 | cv2.imwrite(temp_frame_path, result) 223 | except Exception as exception: 224 | print(exception) 225 | pass 226 | if progress: 227 | progress.update(1) 228 | else: 229 | for temp_frame_path in temp_frame_paths: 230 | temp_frame = cv2.imread(temp_frame_path) 231 | try: 232 | result = process_frame_v2(temp_frame, temp_frame_path) 233 | cv2.imwrite(temp_frame_path, result) 234 | except Exception as exception: 235 | print(exception) 236 | pass 237 | if progress: 238 | progress.update(1) 239 | 240 | 241 | def process_image(source_path: str, target_path: str, output_path: str) -> None: 242 | if not modules.globals.map_faces: 243 | source_face = get_one_face(cv2.imread(source_path)) 244 | target_frame = cv2.imread(target_path) 245 | result = process_frame(source_face, target_frame) 246 | cv2.imwrite(output_path, result) 247 | else: 248 | if modules.globals.many_faces: 249 | update_status( 250 | "Many faces enabled. Using first source image. Progressing...", NAME 251 | ) 252 | target_frame = cv2.imread(output_path) 253 | result = process_frame_v2(target_frame) 254 | cv2.imwrite(output_path, result) 255 | 256 | 257 | def process_video(source_path: str, temp_frame_paths: List[str]) -> None: 258 | if modules.globals.map_faces and modules.globals.many_faces: 259 | update_status( 260 | "Many faces enabled. Using first source image. Progressing...", NAME 261 | ) 262 | modules.processors.frame.core.process_video( 263 | source_path, temp_frame_paths, process_frames 264 | ) 265 | 266 | 267 | def create_lower_mouth_mask( 268 | face: Face, frame: Frame 269 | ) -> (np.ndarray, np.ndarray, tuple, np.ndarray): 270 | mask = np.zeros(frame.shape[:2], dtype=np.uint8) 271 | mouth_cutout = None 272 | landmarks = face.landmark_2d_106 273 | if landmarks is not None: 274 | # 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 275 | lower_lip_order = [ 276 | 65, 277 | 66, 278 | 62, 279 | 70, 280 | 69, 281 | 18, 282 | 19, 283 | 20, 284 | 21, 285 | 22, 286 | 23, 287 | 24, 288 | 0, 289 | 8, 290 | 7, 291 | 6, 292 | 5, 293 | 4, 294 | 3, 295 | 2, 296 | 65, 297 | ] 298 | lower_lip_landmarks = landmarks[lower_lip_order].astype( 299 | np.float32 300 | ) # Use float for precise calculations 301 | 302 | # Calculate the center of the landmarks 303 | center = np.mean(lower_lip_landmarks, axis=0) 304 | 305 | # Expand the landmarks outward 306 | expansion_factor = ( 307 | 1 + modules.globals.mask_down_size 308 | ) # Adjust this for more or less expansion 309 | expanded_landmarks = (lower_lip_landmarks - center) * expansion_factor + center 310 | 311 | # Extend the top lip part 312 | toplip_indices = [ 313 | 20, 314 | 0, 315 | 1, 316 | 2, 317 | 3, 318 | 4, 319 | 5, 320 | ] # Indices for landmarks 2, 65, 66, 62, 70, 69, 18 321 | toplip_extension = ( 322 | modules.globals.mask_size * 0.5 323 | ) # Adjust this factor to control the extension 324 | for idx in toplip_indices: 325 | direction = expanded_landmarks[idx] - center 326 | direction = direction / np.linalg.norm(direction) 327 | expanded_landmarks[idx] += direction * toplip_extension 328 | 329 | # Extend the bottom part (chin area) 330 | chin_indices = [ 331 | 11, 332 | 12, 333 | 13, 334 | 14, 335 | 15, 336 | 16, 337 | ] # Indices for landmarks 21, 22, 23, 24, 0, 8 338 | chin_extension = 2 * 0.2 # Adjust this factor to control the extension 339 | for idx in chin_indices: 340 | expanded_landmarks[idx][1] += ( 341 | expanded_landmarks[idx][1] - center[1] 342 | ) * chin_extension 343 | 344 | # Convert back to integer coordinates 345 | expanded_landmarks = expanded_landmarks.astype(np.int32) 346 | 347 | # Calculate bounding box for the expanded lower mouth 348 | min_x, min_y = np.min(expanded_landmarks, axis=0) 349 | max_x, max_y = np.max(expanded_landmarks, axis=0) 350 | 351 | # Add some padding to the bounding box 352 | padding = int((max_x - min_x) * 0.1) # 10% padding 353 | min_x = max(0, min_x - padding) 354 | min_y = max(0, min_y - padding) 355 | max_x = min(frame.shape[1], max_x + padding) 356 | max_y = min(frame.shape[0], max_y + padding) 357 | 358 | # Ensure the bounding box dimensions are valid 359 | if max_x <= min_x or max_y <= min_y: 360 | if (max_x - min_x) <= 1: 361 | max_x = min_x + 1 362 | if (max_y - min_y) <= 1: 363 | max_y = min_y + 1 364 | 365 | # Create the mask 366 | mask_roi = np.zeros((max_y - min_y, max_x - min_x), dtype=np.uint8) 367 | cv2.fillPoly(mask_roi, [expanded_landmarks - [min_x, min_y]], 255) 368 | 369 | # Apply Gaussian blur to soften the mask edges 370 | mask_roi = cv2.GaussianBlur(mask_roi, (15, 15), 5) 371 | 372 | # Place the mask ROI in the full-sized mask 373 | mask[min_y:max_y, min_x:max_x] = mask_roi 374 | 375 | # Extract the masked area from the frame 376 | mouth_cutout = frame[min_y:max_y, min_x:max_x].copy() 377 | 378 | # Return the expanded lower lip polygon in original frame coordinates 379 | lower_lip_polygon = expanded_landmarks 380 | 381 | return mask, mouth_cutout, (min_x, min_y, max_x, max_y), lower_lip_polygon 382 | 383 | 384 | def draw_mouth_mask_visualization( 385 | frame: Frame, face: Face, mouth_mask_data: tuple 386 | ) -> Frame: 387 | landmarks = face.landmark_2d_106 388 | if landmarks is not None and mouth_mask_data is not None: 389 | mask, mouth_cutout, (min_x, min_y, max_x, max_y), lower_lip_polygon = ( 390 | mouth_mask_data 391 | ) 392 | 393 | vis_frame = frame.copy() 394 | 395 | # Ensure coordinates are within frame bounds 396 | height, width = vis_frame.shape[:2] 397 | min_x, min_y = max(0, min_x), max(0, min_y) 398 | max_x, max_y = min(width, max_x), min(height, max_y) 399 | 400 | # Adjust mask to match the region size 401 | mask_region = mask[0 : max_y - min_y, 0 : max_x - min_x] 402 | 403 | # Remove the color mask overlay 404 | # color_mask = cv2.applyColorMap((mask_region * 255).astype(np.uint8), cv2.COLORMAP_JET) 405 | 406 | # Ensure shapes match before blending 407 | vis_region = vis_frame[min_y:max_y, min_x:max_x] 408 | # Remove blending with color_mask 409 | # if vis_region.shape[:2] == color_mask.shape[:2]: 410 | # blended = cv2.addWeighted(vis_region, 0.7, color_mask, 0.3, 0) 411 | # vis_frame[min_y:max_y, min_x:max_x] = blended 412 | 413 | # Draw the lower lip polygon 414 | cv2.polylines(vis_frame, [lower_lip_polygon], True, (0, 255, 0), 2) 415 | 416 | # Remove the red box 417 | # cv2.rectangle(vis_frame, (min_x, min_y), (max_x, max_y), (0, 0, 255), 2) 418 | 419 | # Visualize the feathered mask 420 | feather_amount = max( 421 | 1, 422 | min( 423 | 30, 424 | (max_x - min_x) // modules.globals.mask_feather_ratio, 425 | (max_y - min_y) // modules.globals.mask_feather_ratio, 426 | ), 427 | ) 428 | # Ensure kernel size is odd 429 | kernel_size = 2 * feather_amount + 1 430 | feathered_mask = cv2.GaussianBlur( 431 | mask_region.astype(float), (kernel_size, kernel_size), 0 432 | ) 433 | feathered_mask = (feathered_mask / feathered_mask.max() * 255).astype(np.uint8) 434 | # Remove the feathered mask color overlay 435 | # color_feathered_mask = cv2.applyColorMap(feathered_mask, cv2.COLORMAP_VIRIDIS) 436 | 437 | # Ensure shapes match before blending feathered mask 438 | # if vis_region.shape == color_feathered_mask.shape: 439 | # blended_feathered = cv2.addWeighted(vis_region, 0.7, color_feathered_mask, 0.3, 0) 440 | # vis_frame[min_y:max_y, min_x:max_x] = blended_feathered 441 | 442 | # Add labels 443 | cv2.putText( 444 | vis_frame, 445 | "Lower Mouth Mask", 446 | (min_x, min_y - 10), 447 | cv2.FONT_HERSHEY_SIMPLEX, 448 | 0.5, 449 | (255, 255, 255), 450 | 1, 451 | ) 452 | cv2.putText( 453 | vis_frame, 454 | "Feathered Mask", 455 | (min_x, max_y + 20), 456 | cv2.FONT_HERSHEY_SIMPLEX, 457 | 0.5, 458 | (255, 255, 255), 459 | 1, 460 | ) 461 | 462 | return vis_frame 463 | return frame 464 | 465 | 466 | def apply_mouth_area( 467 | frame: np.ndarray, 468 | mouth_cutout: np.ndarray, 469 | mouth_box: tuple, 470 | face_mask: np.ndarray, 471 | mouth_polygon: np.ndarray, 472 | ) -> np.ndarray: 473 | min_x, min_y, max_x, max_y = mouth_box 474 | box_width = max_x - min_x 475 | box_height = max_y - min_y 476 | 477 | if ( 478 | mouth_cutout is None 479 | or box_width is None 480 | or box_height is None 481 | or face_mask is None 482 | or mouth_polygon is None 483 | ): 484 | return frame 485 | 486 | try: 487 | resized_mouth_cutout = cv2.resize(mouth_cutout, (box_width, box_height)) 488 | roi = frame[min_y:max_y, min_x:max_x] 489 | 490 | if roi.shape != resized_mouth_cutout.shape: 491 | resized_mouth_cutout = cv2.resize( 492 | resized_mouth_cutout, (roi.shape[1], roi.shape[0]) 493 | ) 494 | 495 | color_corrected_mouth = apply_color_transfer(resized_mouth_cutout, roi) 496 | 497 | # Use the provided mouth polygon to create the mask 498 | polygon_mask = np.zeros(roi.shape[:2], dtype=np.uint8) 499 | adjusted_polygon = mouth_polygon - [min_x, min_y] 500 | cv2.fillPoly(polygon_mask, [adjusted_polygon], 255) 501 | 502 | # Apply feathering to the polygon mask 503 | feather_amount = min( 504 | 30, 505 | box_width // modules.globals.mask_feather_ratio, 506 | box_height // modules.globals.mask_feather_ratio, 507 | ) 508 | feathered_mask = cv2.GaussianBlur( 509 | polygon_mask.astype(float), (0, 0), feather_amount 510 | ) 511 | feathered_mask = feathered_mask / feathered_mask.max() 512 | 513 | face_mask_roi = face_mask[min_y:max_y, min_x:max_x] 514 | combined_mask = feathered_mask * (face_mask_roi / 255.0) 515 | 516 | combined_mask = combined_mask[:, :, np.newaxis] 517 | blended = ( 518 | color_corrected_mouth * combined_mask + roi * (1 - combined_mask) 519 | ).astype(np.uint8) 520 | 521 | # Apply face mask to blended result 522 | face_mask_3channel = ( 523 | np.repeat(face_mask_roi[:, :, np.newaxis], 3, axis=2) / 255.0 524 | ) 525 | final_blend = blended * face_mask_3channel + roi * (1 - face_mask_3channel) 526 | 527 | frame[min_y:max_y, min_x:max_x] = final_blend.astype(np.uint8) 528 | except Exception as e: 529 | pass 530 | 531 | return frame 532 | 533 | 534 | def create_face_mask(face: Face, frame: Frame) -> np.ndarray: 535 | mask = np.zeros(frame.shape[:2], dtype=np.uint8) 536 | landmarks = face.landmark_2d_106 537 | if landmarks is not None: 538 | # Convert landmarks to int32 539 | landmarks = landmarks.astype(np.int32) 540 | 541 | # Extract facial features 542 | right_side_face = landmarks[0:16] 543 | left_side_face = landmarks[17:32] 544 | right_eye = landmarks[33:42] 545 | right_eye_brow = landmarks[43:51] 546 | left_eye = landmarks[87:96] 547 | left_eye_brow = landmarks[97:105] 548 | 549 | # Calculate forehead extension 550 | right_eyebrow_top = np.min(right_eye_brow[:, 1]) 551 | left_eyebrow_top = np.min(left_eye_brow[:, 1]) 552 | eyebrow_top = min(right_eyebrow_top, left_eyebrow_top) 553 | 554 | face_top = np.min([right_side_face[0, 1], left_side_face[-1, 1]]) 555 | forehead_height = face_top - eyebrow_top 556 | extended_forehead_height = int(forehead_height * 5.0) # Extend by 50% 557 | 558 | # Create forehead points 559 | forehead_left = right_side_face[0].copy() 560 | forehead_right = left_side_face[-1].copy() 561 | forehead_left[1] -= extended_forehead_height 562 | forehead_right[1] -= extended_forehead_height 563 | 564 | # Combine all points to create the face outline 565 | face_outline = np.vstack( 566 | [ 567 | [forehead_left], 568 | right_side_face, 569 | left_side_face[ 570 | ::-1 571 | ], # Reverse left side to create a continuous outline 572 | [forehead_right], 573 | ] 574 | ) 575 | 576 | # Calculate padding 577 | padding = int( 578 | np.linalg.norm(right_side_face[0] - left_side_face[-1]) * 0.05 579 | ) # 5% of face width 580 | 581 | # Create a slightly larger convex hull for padding 582 | hull = cv2.convexHull(face_outline) 583 | hull_padded = [] 584 | for point in hull: 585 | x, y = point[0] 586 | center = np.mean(face_outline, axis=0) 587 | direction = np.array([x, y]) - center 588 | direction = direction / np.linalg.norm(direction) 589 | padded_point = np.array([x, y]) + direction * padding 590 | hull_padded.append(padded_point) 591 | 592 | hull_padded = np.array(hull_padded, dtype=np.int32) 593 | 594 | # Fill the padded convex hull 595 | cv2.fillConvexPoly(mask, hull_padded, 255) 596 | 597 | # Smooth the mask edges 598 | mask = cv2.GaussianBlur(mask, (5, 5), 3) 599 | 600 | return mask 601 | 602 | 603 | def apply_color_transfer(source, target): 604 | """ 605 | Apply color transfer from target to source image 606 | """ 607 | source = cv2.cvtColor(source, cv2.COLOR_BGR2LAB).astype("float32") 608 | target = cv2.cvtColor(target, cv2.COLOR_BGR2LAB).astype("float32") 609 | 610 | source_mean, source_std = cv2.meanStdDev(source) 611 | target_mean, target_std = cv2.meanStdDev(target) 612 | 613 | # Reshape mean and std to be broadcastable 614 | source_mean = source_mean.reshape(1, 1, 3) 615 | source_std = source_std.reshape(1, 1, 3) 616 | target_mean = target_mean.reshape(1, 1, 3) 617 | target_std = target_std.reshape(1, 1, 3) 618 | 619 | # Perform the color transfer 620 | source = (source - source_mean) * (target_std / source_std) + target_mean 621 | 622 | return cv2.cvtColor(np.clip(source, 0, 255).astype("uint8"), cv2.COLOR_LAB2BGR) 623 | -------------------------------------------------------------------------------- /modules/typing.py: -------------------------------------------------------------------------------- 1 | from typing import Any 2 | 3 | from insightface.app.common import Face 4 | import numpy 5 | 6 | Face = Face 7 | Frame = numpy.ndarray[Any, Any] 8 | -------------------------------------------------------------------------------- /modules/ui.json: -------------------------------------------------------------------------------- 1 | { 2 | "CTk": { 3 | "fg_color": ["gray95", "gray10"] 4 | }, 5 | "CTkToplevel": { 6 | "fg_color": ["gray95", "gray10"] 7 | }, 8 | "CTkFrame": { 9 | "corner_radius": 0, 10 | "border_width": 0, 11 | "fg_color": ["gray90", "gray13"], 12 | "top_fg_color": ["gray85", "gray16"], 13 | "border_color": ["gray65", "gray28"] 14 | }, 15 | "CTkButton": { 16 | "corner_radius": 0, 17 | "border_width": 0, 18 | "fg_color": ["#2aa666", "#1f538d"], 19 | "hover_color": ["#3cb666", "#14375e"], 20 | "border_color": ["#3e4a40", "#949A9F"], 21 | "text_color": ["#f3faf6", "#f3faf6"], 22 | "text_color_disabled": ["gray74", "gray60"] 23 | }, 24 | "CTkLabel": { 25 | "corner_radius": 0, 26 | "fg_color": "transparent", 27 | "text_color": ["gray14", "gray84"] 28 | }, 29 | "CTkEntry": { 30 | "corner_radius": 0, 31 | "border_width": 2, 32 | "fg_color": ["#F9F9FA", "#343638"], 33 | "border_color": ["#979DA2", "#565B5E"], 34 | "text_color": ["gray14", "gray84"], 35 | "placeholder_text_color": ["gray52", "gray62"] 36 | }, 37 | "CTkCheckbox": { 38 | "corner_radius": 0, 39 | "border_width": 3, 40 | "fg_color": ["#2aa666", "#1f538d"], 41 | "border_color": ["#3e4a40", "#949A9F"], 42 | "hover_color": ["#3cb666", "#14375e"], 43 | "checkmark_color": ["#f3faf6", "gray90"], 44 | "text_color": ["gray14", "gray84"], 45 | "text_color_disabled": ["gray60", "gray45"] 46 | }, 47 | "CTkSwitch": { 48 | "corner_radius": 1000, 49 | "border_width": 3, 50 | "button_length": 0, 51 | "fg_color": ["#939BA2", "#4A4D50"], 52 | "progress_color": ["#2aa666", "#1f538d"], 53 | "button_color": ["gray36", "#D5D9DE"], 54 | "button_hover_color": ["gray20", "gray100"], 55 | "text_color": ["gray14", "gray84"], 56 | "text_color_disabled": ["gray60", "gray45"] 57 | }, 58 | "CTkRadiobutton": { 59 | "corner_radius": 1000, 60 | "border_width_checked": 6, 61 | "border_width_unchecked": 3, 62 | "fg_color": ["#2aa666", "#1f538d"], 63 | "border_color": ["#3e4a40", "#949A9F"], 64 | "hover_color": ["#3cb666", "#14375e"], 65 | "text_color": ["gray14", "gray84"], 66 | "text_color_disabled": ["gray60", "gray45"] 67 | }, 68 | "CTkProgressBar": { 69 | "corner_radius": 1000, 70 | "border_width": 0, 71 | "fg_color": ["#939BA2", "#4A4D50"], 72 | "progress_color": ["#2aa666", "#1f538d"], 73 | "border_color": ["gray", "gray"] 74 | }, 75 | "CTkSlider": { 76 | "corner_radius": 1000, 77 | "button_corner_radius": 1000, 78 | "border_width": 6, 79 | "button_length": 0, 80 | "fg_color": ["#939BA2", "#4A4D50"], 81 | "progress_color": ["gray40", "#AAB0B5"], 82 | "button_color": ["#2aa666", "#1f538d"], 83 | "button_hover_color": ["#3cb666", "#14375e"] 84 | }, 85 | "CTkOptionMenu": { 86 | "corner_radius": 0, 87 | "fg_color": ["#2aa666", "#1f538d"], 88 | "button_color": ["#3cb666", "#14375e"], 89 | "button_hover_color": ["#234567", "#1e2c40"], 90 | "text_color": ["#f3faf6", "#f3faf6"], 91 | "text_color_disabled": ["gray74", "gray60"] 92 | }, 93 | "CTkComboBox": { 94 | "corner_radius": 0, 95 | "border_width": 2, 96 | "fg_color": ["#F9F9FA", "#343638"], 97 | "border_color": ["#979DA2", "#565B5E"], 98 | "button_color": ["#979DA2", "#565B5E"], 99 | "button_hover_color": ["#6E7174", "#7A848D"], 100 | "text_color": ["gray14", "gray84"], 101 | "text_color_disabled": ["gray50", "gray45"] 102 | }, 103 | "CTkScrollbar": { 104 | "corner_radius": 1000, 105 | "border_spacing": 4, 106 | "fg_color": "transparent", 107 | "button_color": ["gray55", "gray41"], 108 | "button_hover_color": ["gray40", "gray53"] 109 | }, 110 | "CTkSegmentedButton": { 111 | "corner_radius": 0, 112 | "border_width": 2, 113 | "fg_color": ["#979DA2", "gray29"], 114 | "selected_color": ["#2aa666", "#1f538d"], 115 | "selected_hover_color": ["#3cb666", "#14375e"], 116 | "unselected_color": ["#979DA2", "gray29"], 117 | "unselected_hover_color": ["gray70", "gray41"], 118 | "text_color": ["#f3faf6", "#f3faf6"], 119 | "text_color_disabled": ["gray74", "gray60"] 120 | }, 121 | "CTkTextbox": { 122 | "corner_radius": 0, 123 | "border_width": 0, 124 | "fg_color": ["gray100", "gray20"], 125 | "border_color": ["#979DA2", "#565B5E"], 126 | "text_color": ["gray14", "gray84"], 127 | "scrollbar_button_color": ["gray55", "gray41"], 128 | "scrollbar_button_hover_color": ["gray40", "gray53"] 129 | }, 130 | "CTkScrollableFrame": { 131 | "label_fg_color": ["gray80", "gray21"] 132 | }, 133 | "DropdownMenu": { 134 | "fg_color": ["gray90", "gray20"], 135 | "hover_color": ["gray75", "gray28"], 136 | "text_color": ["gray14", "gray84"] 137 | }, 138 | "CTkFont": { 139 | "macOS": { 140 | "family": "Avenir", 141 | "size": 18, 142 | "weight": "normal" 143 | }, 144 | "Windows": { 145 | "family": "Corbel", 146 | "size": 18, 147 | "weight": "normal" 148 | }, 149 | "Linux": { 150 | "family": "Montserrat", 151 | "size": 18, 152 | "weight": "normal" 153 | } 154 | }, 155 | "URL": { 156 | "text_color": ["gray74", "gray60"] 157 | } 158 | } 159 | -------------------------------------------------------------------------------- /modules/ui.py: -------------------------------------------------------------------------------- 1 | import os 2 | import webbrowser 3 | import customtkinter as ctk 4 | from typing import Callable, Tuple 5 | import cv2 6 | from cv2_enumerate_cameras import enumerate_cameras # Add this import 7 | from PIL import Image, ImageOps 8 | import time 9 | import json 10 | import modules.globals 11 | import modules.metadata 12 | from modules.face_analyser import ( 13 | get_one_face, 14 | get_unique_faces_from_target_image, 15 | get_unique_faces_from_target_video, 16 | add_blank_map, 17 | has_valid_map, 18 | simplify_maps, 19 | ) 20 | from modules.capturer import get_video_frame, get_video_frame_total 21 | from modules.processors.frame.core import get_frame_processors_modules 22 | from modules.utilities import ( 23 | is_image, 24 | is_video, 25 | resolve_relative_path, 26 | has_image_extension, 27 | ) 28 | from modules.video_capture import VideoCapturer 29 | from modules.gettext import LanguageManager 30 | import platform 31 | 32 | if platform.system() == "Windows": 33 | from pygrabber.dshow_graph import FilterGraph 34 | 35 | ROOT = None 36 | POPUP = None 37 | POPUP_LIVE = None 38 | ROOT_HEIGHT = 700 39 | ROOT_WIDTH = 600 40 | 41 | PREVIEW = None 42 | PREVIEW_MAX_HEIGHT = 700 43 | PREVIEW_MAX_WIDTH = 1200 44 | PREVIEW_DEFAULT_WIDTH = 960 45 | PREVIEW_DEFAULT_HEIGHT = 540 46 | 47 | POPUP_WIDTH = 750 48 | POPUP_HEIGHT = 810 49 | POPUP_SCROLL_WIDTH = (740,) 50 | POPUP_SCROLL_HEIGHT = 700 51 | 52 | POPUP_LIVE_WIDTH = 900 53 | POPUP_LIVE_HEIGHT = 820 54 | POPUP_LIVE_SCROLL_WIDTH = (890,) 55 | POPUP_LIVE_SCROLL_HEIGHT = 700 56 | 57 | MAPPER_PREVIEW_MAX_HEIGHT = 100 58 | MAPPER_PREVIEW_MAX_WIDTH = 100 59 | 60 | DEFAULT_BUTTON_WIDTH = 200 61 | DEFAULT_BUTTON_HEIGHT = 40 62 | 63 | RECENT_DIRECTORY_SOURCE = None 64 | RECENT_DIRECTORY_TARGET = None 65 | RECENT_DIRECTORY_OUTPUT = None 66 | 67 | _ = None 68 | preview_label = None 69 | preview_slider = None 70 | source_label = None 71 | target_label = None 72 | status_label = None 73 | popup_status_label = None 74 | popup_status_label_live = None 75 | source_label_dict = {} 76 | source_label_dict_live = {} 77 | target_label_dict_live = {} 78 | 79 | img_ft, vid_ft = modules.globals.file_types 80 | 81 | 82 | def init(start: Callable[[], None], destroy: Callable[[], None], lang: str) -> ctk.CTk: 83 | global ROOT, PREVIEW, _ 84 | 85 | lang_manager = LanguageManager(lang) 86 | _ = lang_manager._ 87 | ROOT = create_root(start, destroy) 88 | PREVIEW = create_preview(ROOT) 89 | 90 | return ROOT 91 | 92 | 93 | def save_switch_states(): 94 | switch_states = { 95 | "keep_fps": modules.globals.keep_fps, 96 | "keep_audio": modules.globals.keep_audio, 97 | "keep_frames": modules.globals.keep_frames, 98 | "many_faces": modules.globals.many_faces, 99 | "map_faces": modules.globals.map_faces, 100 | "color_correction": modules.globals.color_correction, 101 | "nsfw_filter": modules.globals.nsfw_filter, 102 | "live_mirror": modules.globals.live_mirror, 103 | "live_resizable": modules.globals.live_resizable, 104 | "fp_ui": modules.globals.fp_ui, 105 | "show_fps": modules.globals.show_fps, 106 | "mouth_mask": modules.globals.mouth_mask, 107 | "show_mouth_mask_box": modules.globals.show_mouth_mask_box, 108 | } 109 | with open("switch_states.json", "w") as f: 110 | json.dump(switch_states, f) 111 | 112 | 113 | def load_switch_states(): 114 | try: 115 | with open("switch_states.json", "r") as f: 116 | switch_states = json.load(f) 117 | modules.globals.keep_fps = switch_states.get("keep_fps", True) 118 | modules.globals.keep_audio = switch_states.get("keep_audio", True) 119 | modules.globals.keep_frames = switch_states.get("keep_frames", False) 120 | modules.globals.many_faces = switch_states.get("many_faces", False) 121 | modules.globals.map_faces = switch_states.get("map_faces", False) 122 | modules.globals.color_correction = switch_states.get("color_correction", False) 123 | modules.globals.nsfw_filter = switch_states.get("nsfw_filter", False) 124 | modules.globals.live_mirror = switch_states.get("live_mirror", False) 125 | modules.globals.live_resizable = switch_states.get("live_resizable", False) 126 | modules.globals.fp_ui = switch_states.get("fp_ui", {"face_enhancer": False}) 127 | modules.globals.show_fps = switch_states.get("show_fps", False) 128 | modules.globals.mouth_mask = switch_states.get("mouth_mask", False) 129 | modules.globals.show_mouth_mask_box = switch_states.get( 130 | "show_mouth_mask_box", False 131 | ) 132 | except FileNotFoundError: 133 | # If the file doesn't exist, use default values 134 | pass 135 | 136 | 137 | def create_root(start: Callable[[], None], destroy: Callable[[], None]) -> ctk.CTk: 138 | global source_label, target_label, status_label, show_fps_switch 139 | 140 | load_switch_states() 141 | 142 | ctk.deactivate_automatic_dpi_awareness() 143 | ctk.set_appearance_mode("system") 144 | ctk.set_default_color_theme(resolve_relative_path("ui.json")) 145 | 146 | root = ctk.CTk() 147 | root.minsize(ROOT_WIDTH, ROOT_HEIGHT) 148 | root.title( 149 | f"{modules.metadata.name} {modules.metadata.version} {modules.metadata.edition}" 150 | ) 151 | root.configure() 152 | root.protocol("WM_DELETE_WINDOW", lambda: destroy()) 153 | 154 | source_label = ctk.CTkLabel(root, text=None) 155 | source_label.place(relx=0.1, rely=0.1, relwidth=0.3, relheight=0.25) 156 | 157 | target_label = ctk.CTkLabel(root, text=None) 158 | target_label.place(relx=0.6, rely=0.1, relwidth=0.3, relheight=0.25) 159 | 160 | select_face_button = ctk.CTkButton( 161 | root, text=_("Select a face"), cursor="hand2", command=lambda: select_source_path() 162 | ) 163 | select_face_button.place(relx=0.1, rely=0.4, relwidth=0.3, relheight=0.1) 164 | 165 | swap_faces_button = ctk.CTkButton( 166 | root, text="↔", cursor="hand2", command=lambda: swap_faces_paths() 167 | ) 168 | swap_faces_button.place(relx=0.45, rely=0.4, relwidth=0.1, relheight=0.1) 169 | 170 | select_target_button = ctk.CTkButton( 171 | root, 172 | text=_("Select a target"), 173 | cursor="hand2", 174 | command=lambda: select_target_path(), 175 | ) 176 | select_target_button.place(relx=0.6, rely=0.4, relwidth=0.3, relheight=0.1) 177 | 178 | keep_fps_value = ctk.BooleanVar(value=modules.globals.keep_fps) 179 | keep_fps_checkbox = ctk.CTkSwitch( 180 | root, 181 | text=_("Keep fps"), 182 | variable=keep_fps_value, 183 | cursor="hand2", 184 | command=lambda: ( 185 | setattr(modules.globals, "keep_fps", keep_fps_value.get()), 186 | save_switch_states(), 187 | ), 188 | ) 189 | keep_fps_checkbox.place(relx=0.1, rely=0.6) 190 | 191 | keep_frames_value = ctk.BooleanVar(value=modules.globals.keep_frames) 192 | keep_frames_switch = ctk.CTkSwitch( 193 | root, 194 | text=_("Keep frames"), 195 | variable=keep_frames_value, 196 | cursor="hand2", 197 | command=lambda: ( 198 | setattr(modules.globals, "keep_frames", keep_frames_value.get()), 199 | save_switch_states(), 200 | ), 201 | ) 202 | keep_frames_switch.place(relx=0.1, rely=0.65) 203 | 204 | enhancer_value = ctk.BooleanVar(value=modules.globals.fp_ui["face_enhancer"]) 205 | enhancer_switch = ctk.CTkSwitch( 206 | root, 207 | text=_("Face Enhancer"), 208 | variable=enhancer_value, 209 | cursor="hand2", 210 | command=lambda: ( 211 | update_tumbler("face_enhancer", enhancer_value.get()), 212 | save_switch_states(), 213 | ), 214 | ) 215 | enhancer_switch.place(relx=0.1, rely=0.7) 216 | 217 | keep_audio_value = ctk.BooleanVar(value=modules.globals.keep_audio) 218 | keep_audio_switch = ctk.CTkSwitch( 219 | root, 220 | text=_("Keep audio"), 221 | variable=keep_audio_value, 222 | cursor="hand2", 223 | command=lambda: ( 224 | setattr(modules.globals, "keep_audio", keep_audio_value.get()), 225 | save_switch_states(), 226 | ), 227 | ) 228 | keep_audio_switch.place(relx=0.6, rely=0.6) 229 | 230 | many_faces_value = ctk.BooleanVar(value=modules.globals.many_faces) 231 | many_faces_switch = ctk.CTkSwitch( 232 | root, 233 | text=_("Many faces"), 234 | variable=many_faces_value, 235 | cursor="hand2", 236 | command=lambda: ( 237 | setattr(modules.globals, "many_faces", many_faces_value.get()), 238 | save_switch_states(), 239 | ), 240 | ) 241 | many_faces_switch.place(relx=0.6, rely=0.65) 242 | 243 | color_correction_value = ctk.BooleanVar(value=modules.globals.color_correction) 244 | color_correction_switch = ctk.CTkSwitch( 245 | root, 246 | text=_("Fix Blueish Cam"), 247 | variable=color_correction_value, 248 | cursor="hand2", 249 | command=lambda: ( 250 | setattr(modules.globals, "color_correction", color_correction_value.get()), 251 | save_switch_states(), 252 | ), 253 | ) 254 | color_correction_switch.place(relx=0.6, rely=0.70) 255 | 256 | # nsfw_value = ctk.BooleanVar(value=modules.globals.nsfw_filter) 257 | # nsfw_switch = ctk.CTkSwitch(root, text='NSFW filter', variable=nsfw_value, cursor='hand2', command=lambda: setattr(modules.globals, 'nsfw_filter', nsfw_value.get())) 258 | # nsfw_switch.place(relx=0.6, rely=0.7) 259 | 260 | map_faces = ctk.BooleanVar(value=modules.globals.map_faces) 261 | map_faces_switch = ctk.CTkSwitch( 262 | root, 263 | text=_("Map faces"), 264 | variable=map_faces, 265 | cursor="hand2", 266 | command=lambda: ( 267 | setattr(modules.globals, "map_faces", map_faces.get()), 268 | save_switch_states(), 269 | close_mapper_window() if not map_faces.get() else None 270 | ), 271 | ) 272 | map_faces_switch.place(relx=0.1, rely=0.75) 273 | 274 | show_fps_value = ctk.BooleanVar(value=modules.globals.show_fps) 275 | show_fps_switch = ctk.CTkSwitch( 276 | root, 277 | text=_("Show FPS"), 278 | variable=show_fps_value, 279 | cursor="hand2", 280 | command=lambda: ( 281 | setattr(modules.globals, "show_fps", show_fps_value.get()), 282 | save_switch_states(), 283 | ), 284 | ) 285 | show_fps_switch.place(relx=0.6, rely=0.75) 286 | 287 | mouth_mask_var = ctk.BooleanVar(value=modules.globals.mouth_mask) 288 | mouth_mask_switch = ctk.CTkSwitch( 289 | root, 290 | text=_("Mouth Mask"), 291 | variable=mouth_mask_var, 292 | cursor="hand2", 293 | command=lambda: setattr(modules.globals, "mouth_mask", mouth_mask_var.get()), 294 | ) 295 | mouth_mask_switch.place(relx=0.1, rely=0.55) 296 | 297 | show_mouth_mask_box_var = ctk.BooleanVar(value=modules.globals.show_mouth_mask_box) 298 | show_mouth_mask_box_switch = ctk.CTkSwitch( 299 | root, 300 | text=_("Show Mouth Mask Box"), 301 | variable=show_mouth_mask_box_var, 302 | cursor="hand2", 303 | command=lambda: setattr( 304 | modules.globals, "show_mouth_mask_box", show_mouth_mask_box_var.get() 305 | ), 306 | ) 307 | show_mouth_mask_box_switch.place(relx=0.6, rely=0.55) 308 | 309 | start_button = ctk.CTkButton( 310 | root, text=_("Start"), cursor="hand2", command=lambda: analyze_target(start, root) 311 | ) 312 | start_button.place(relx=0.15, rely=0.80, relwidth=0.2, relheight=0.05) 313 | 314 | stop_button = ctk.CTkButton( 315 | root, text=_("Destroy"), cursor="hand2", command=lambda: destroy() 316 | ) 317 | stop_button.place(relx=0.4, rely=0.80, relwidth=0.2, relheight=0.05) 318 | 319 | preview_button = ctk.CTkButton( 320 | root, text=_("Preview"), cursor="hand2", command=lambda: toggle_preview() 321 | ) 322 | preview_button.place(relx=0.65, rely=0.80, relwidth=0.2, relheight=0.05) 323 | 324 | # --- Camera Selection --- 325 | camera_label = ctk.CTkLabel(root, text=_("Select Camera:")) 326 | camera_label.place(relx=0.1, rely=0.86, relwidth=0.2, relheight=0.05) 327 | 328 | available_cameras = get_available_cameras() 329 | camera_indices, camera_names = available_cameras 330 | 331 | if not camera_names or camera_names[0] == "No cameras found": 332 | camera_variable = ctk.StringVar(value="No cameras found") 333 | camera_optionmenu = ctk.CTkOptionMenu( 334 | root, 335 | variable=camera_variable, 336 | values=["No cameras found"], 337 | state="disabled", 338 | ) 339 | else: 340 | camera_variable = ctk.StringVar(value=camera_names[0]) 341 | camera_optionmenu = ctk.CTkOptionMenu( 342 | root, variable=camera_variable, values=camera_names 343 | ) 344 | 345 | camera_optionmenu.place(relx=0.35, rely=0.86, relwidth=0.25, relheight=0.05) 346 | 347 | live_button = ctk.CTkButton( 348 | root, 349 | text=_("Live"), 350 | cursor="hand2", 351 | command=lambda: webcam_preview( 352 | root, 353 | ( 354 | camera_indices[camera_names.index(camera_variable.get())] 355 | if camera_names and camera_names[0] != "No cameras found" 356 | else None 357 | ), 358 | ), 359 | state=( 360 | "normal" 361 | if camera_names and camera_names[0] != "No cameras found" 362 | else "disabled" 363 | ), 364 | ) 365 | live_button.place(relx=0.65, rely=0.86, relwidth=0.2, relheight=0.05) 366 | # --- End Camera Selection --- 367 | 368 | status_label = ctk.CTkLabel(root, text=None, justify="center") 369 | status_label.place(relx=0.1, rely=0.9, relwidth=0.8) 370 | 371 | donate_label = ctk.CTkLabel( 372 | root, text="Deep Live Cam", justify="center", cursor="hand2" 373 | ) 374 | donate_label.place(relx=0.1, rely=0.95, relwidth=0.8) 375 | donate_label.configure( 376 | text_color=ctk.ThemeManager.theme.get("URL").get("text_color") 377 | ) 378 | donate_label.bind( 379 | "