├── LICENSE ├── README.md ├── Results.ipynb ├── data ├── attributes.npy ├── df.csv ├── df_classes.csv └── features_resnet101_10crop.npy ├── dataset.py └── flexible_triplet_loss.py /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 | # Modeling Inter and Intra-Class Relations in the Triplet Loss for Zero-Shot Learning 2 | 3 | This code allows to reprodure the results published in the [ICCV'19 article](http://openaccess.thecvf.com/content_ICCV_2019/papers/Le_Cacheux_Modeling_Inter_and_Intra-Class_Relations_in_the_Triplet_Loss_for_ICCV_2019_paper.pdf) and its [supplementary material](http://openaccess.thecvf.com/content_ICCV_2019/supplemental/Le_Cacheux_Modeling_Inter_and_ICCV_2019_supplemental.pdf). 4 | 5 | All the material is contained in a unique jupyter notebook that you run with: 6 | 7 | ``` 8 | jupyter notebook Results.ipynb 9 | ``` 10 | 11 | Dependencies are: 12 | * python 2.7.9 13 | * torch 1.1.0 14 | * sklearn 0.19.2 15 | * numpy 1.14.3 16 | * pandas 0.23.0 17 | * matplotlib 2.2.2 18 | * scipy 1.0.0 19 | * jupyter 20 | 21 | ## Get the CUB dataset 22 | To visualize the images, you need to download the [CUB 200 2011 dataset](http://www.vision.caltech.edu/visipedia-data/CUB-200-2011/CUB_200_2011.tgz). The corresponding [project page](http://www.vision.caltech.edu/visipedia/CUB-200-2011.html) gives the reference to cite if you use it in your work. 23 | 24 | You need to decompress the dataset into *CUB_200_2011/* then change the path in *data/df.cvs* accordingly, e.g: 25 | 26 | ``` perl 27 | perl -i.old -p -e 's#/scratch_global/yannick#'$PWD'#' data/df.csv 28 | ``` 29 | 30 | ## Citation 31 | Please cite the following article if you use this code in your work: 32 | 33 | Y. Le Cacheux, H. Le Borgne and M. Crucianu. Modeling Inter and Intra-Class Relations in the Triplet Loss for Zero-Shot Learning. In *Proceedings of the IEEE International Conference on Computer Vision, ICCV*, Seoul, Korea, Oct. 27 - Nov. 2, 2019 34 | 35 | ``` 36 | @inproceedings{lecacheux2019zsl, 37 | title = {Modeling Inter and Intra-Class Relations in the Triplet Loss for Zero-Shot Learning}, 38 | author = {Le Cacheux, Yannick and Le Borgne, Herv{\'e} and Crucianu, Michel}, 39 | booktitle = {the IEEE International Conference on Computer Vision (ICCV)}, 40 | month = {October}, 41 | series = {ICCV}, 42 | year = {2019} 43 | } 44 | 45 | ``` 46 | -------------------------------------------------------------------------------- /data/attributes.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/yannick-lc/iccv2019-triplet-loss/adac3ab2e9b79f7aba1bb170194d9bae39099bcd/data/attributes.npy -------------------------------------------------------------------------------- /data/df_classes.csv: -------------------------------------------------------------------------------- 1 | class,original_class,class_name,split_ss1,split_ss2,split_ss3,split_ps1,split_ps2,split_ps3 2 | 0,1,001.Black_footed_Albatross,test,test,test,train,val,train 3 | 1,2,002.Laysan_Albatross,val,train,val,train,val,train 4 | 2,3,003.Sooty_Albatross,train,train,train,train,train,train 5 | 3,4,004.Groove_billed_Ani,test,test,test,test,test,test 6 | 4,5,005.Crested_Auklet,train,val,train,val,train,val 7 | 5,6,006.Least_Auklet,test,test,test,train,train,val 8 | 6,7,007.Parakeet_Auklet,train,train,train,val,val,val 9 | 7,8,008.Rhinoceros_Auklet,test,test,test,train,train,val 10 | 8,9,009.Brewer_Blackbird,test,test,test,val,val,val 11 | 9,10,010.Red_winged_Blackbird,train,train,val,train,train,train 12 | 10,11,011.Rusty_Blackbird,train,train,val,train,val,train 13 | 11,12,012.Yellow_headed_Blackbird,val,train,train,test,test,test 14 | 12,13,013.Bobolink,val,train,train,train,train,val 15 | 13,14,014.Indigo_Bunting,test,test,test,train,val,train 16 | 14,15,015.Lazuli_Bunting,train,val,train,train,val,val 17 | 15,16,016.Painted_Bunting,train,val,train,train,train,train 18 | 16,17,017.Cardinal,val,train,val,train,val,train 19 | 17,18,018.Spotted_Catbird,train,train,val,val,val,val 20 | 18,19,019.Gray_Catbird,val,train,train,train,val,train 21 | 19,20,020.Yellow_breasted_Chat,train,val,train,train,train,val 22 | 20,21,021.Eastern_Towhee,val,train,val,train,train,train 23 | 21,22,022.Chuck_will_Widow,val,val,train,train,train,train 24 | 22,23,023.Brandt_Cormorant,test,test,test,test,test,test 25 | 23,24,024.Red_faced_Cormorant,train,val,train,train,train,train 26 | 24,25,025.Pelagic_Cormorant,train,train,train,train,train,train 27 | 25,26,026.Bronzed_Cowbird,val,train,train,test,test,test 28 | 26,27,027.Shiny_Cowbird,val,val,train,train,train,train 29 | 27,28,028.Brown_Creeper,train,train,train,test,test,test 30 | 28,29,029.American_Crow,test,test,test,val,train,train 31 | 29,30,030.Fish_Crow,val,train,val,train,train,val 32 | 30,31,031.Black_billed_Cuckoo,test,test,test,test,test,test 33 | 31,32,032.Mangrove_Cuckoo,train,val,train,train,train,val 34 | 32,33,033.Yellow_billed_Cuckoo,test,test,test,test,test,test 35 | 33,34,034.Gray_crowned_Rosy_Finch,test,test,test,val,val,train 36 | 34,35,035.Purple_Finch,test,test,test,train,val,train 37 | 35,36,036.Northern_Flicker,test,test,test,val,train,train 38 | 36,37,037.Acadian_Flycatcher,test,test,test,train,train,val 39 | 37,38,038.Great_Crested_Flycatcher,test,test,test,train,val,val 40 | 38,39,039.Least_Flycatcher,train,train,val,train,val,val 41 | 39,40,040.Olive_sided_Flycatcher,train,train,train,train,train,train 42 | 40,41,041.Scissor_tailed_Flycatcher,val,train,train,train,train,val 43 | 41,42,042.Vermilion_Flycatcher,train,train,train,train,train,train 44 | 42,43,043.Yellow_bellied_Flycatcher,test,test,test,test,test,test 45 | 43,44,044.Frigatebird,train,train,val,train,val,train 46 | 44,45,045.Northern_Fulmar,val,val,val,test,test,test 47 | 45,46,046.Gadwall,val,val,val,train,train,train 48 | 46,47,047.American_Goldfinch,train,train,train,train,train,train 49 | 47,48,048.European_Goldfinch,train,train,train,val,train,train 50 | 48,49,049.Boat_tailed_Grackle,test,test,test,test,test,test 51 | 49,50,050.Eared_Grebe,train,train,val,train,train,val 52 | 50,51,051.Horned_Grebe,test,test,test,val,train,train 53 | 51,52,052.Pied_billed_Grebe,train,train,train,test,test,test 54 | 52,53,053.Western_Grebe,test,test,test,train,train,train 55 | 53,54,054.Blue_Grosbeak,train,val,train,train,train,train 56 | 54,55,055.Evening_Grosbeak,val,val,train,test,test,test 57 | 55,56,056.Pine_Grosbeak,train,val,train,val,val,val 58 | 56,57,057.Rose_breasted_Grosbeak,train,val,train,train,train,train 59 | 57,58,058.Pigeon_Guillemot,train,val,train,train,val,train 60 | 58,59,059.California_Gull,train,train,val,val,train,train 61 | 59,60,060.Glaucous_winged_Gull,train,train,train,train,val,val 62 | 60,61,061.Heermann_Gull,val,train,train,val,val,train 63 | 61,62,062.Herring_Gull,train,train,train,train,train,train 64 | 62,63,063.Ivory_Gull,val,train,train,val,val,train 65 | 63,64,064.Ring_billed_Gull,val,train,train,val,train,val 66 | 64,65,065.Slaty_backed_Gull,val,train,val,train,train,train 67 | 65,66,066.Western_Gull,test,test,test,val,val,train 68 | 66,67,067.Anna_Hummingbird,val,train,train,train,train,train 69 | 67,68,068.Ruby_throated_Hummingbird,val,train,val,train,train,train 70 | 68,69,069.Rufous_Hummingbird,train,train,train,val,val,train 71 | 69,70,070.Green_Violetear,train,train,val,test,test,test 72 | 70,71,071.Long_tailed_Jaeger,train,train,train,train,train,train 73 | 71,72,072.Pomarine_Jaeger,test,test,test,test,test,test 74 | 72,73,073.Blue_Jay,train,val,train,val,val,train 75 | 73,74,074.Florida_Jay,val,val,train,val,train,train 76 | 74,75,075.Green_Jay,train,train,train,train,train,val 77 | 75,76,076.Dark_eyed_Junco,train,train,val,val,train,train 78 | 76,77,077.Tropical_Kingbird,train,val,train,test,test,test 79 | 77,78,078.Gray_Kingbird,train,train,train,train,val,train 80 | 78,79,079.Belted_Kingfisher,test,test,test,train,train,val 81 | 79,80,080.Green_Kingfisher,val,train,val,val,train,val 82 | 80,81,081.Pied_Kingfisher,train,val,train,train,train,train 83 | 81,82,082.Ringed_Kingfisher,val,train,val,train,train,train 84 | 82,83,083.White_breasted_Kingfisher,test,test,test,val,val,train 85 | 83,84,084.Red_legged_Kittiwake,test,test,test,test,test,test 86 | 84,85,085.Horned_Lark,train,val,train,train,train,train 87 | 85,86,086.Pacific_Loon,test,test,test,val,val,val 88 | 86,87,087.Mallard,train,train,train,test,test,test 89 | 87,88,088.Western_Meadowlark,train,train,train,train,train,val 90 | 88,89,089.Hooded_Merganser,val,val,val,val,train,train 91 | 89,90,090.Red_breasted_Merganser,train,val,val,train,train,val 92 | 90,91,091.Mockingbird,test,test,test,test,test,test 93 | 91,92,092.Nighthawk,train,val,val,train,train,train 94 | 92,93,093.Clark_Nutcracker,train,train,train,train,val,train 95 | 93,94,094.White_breasted_Nuthatch,train,val,train,test,test,test 96 | 94,95,095.Baltimore_Oriole,test,test,test,train,train,train 97 | 95,96,096.Hooded_Oriole,test,test,test,train,val,train 98 | 96,97,097.Orchard_Oriole,val,train,train,test,test,test 99 | 97,98,098.Scott_Oriole,test,test,test,test,test,test 100 | 98,99,099.Ovenbird,train,train,val,train,train,train 101 | 99,100,100.Brown_Pelican,train,val,train,val,train,train 102 | 100,101,101.White_Pelican,test,test,test,train,train,val 103 | 101,102,102.Western_Wood_Pewee,test,test,test,train,val,val 104 | 102,103,103.Sayornis,test,test,test,test,test,test 105 | 103,104,104.American_Pipit,train,train,train,test,test,test 106 | 104,105,105.Whip_poor_Will,val,val,train,train,train,val 107 | 105,106,106.Horned_Puffin,train,train,train,train,val,train 108 | 106,107,107.Common_Raven,train,val,train,train,train,train 109 | 107,108,108.White_necked_Raven,train,val,val,train,train,val 110 | 108,109,109.American_Redstart,val,train,val,val,train,train 111 | 109,110,110.Geococcyx,train,val,train,train,train,val 112 | 110,111,111.Loggerhead_Shrike,val,train,train,test,test,test 113 | 111,112,112.Great_Grey_Shrike,test,test,test,train,train,val 114 | 112,113,113.Baird_Sparrow,train,val,val,test,test,test 115 | 113,114,114.Black_throated_Sparrow,test,test,test,val,train,val 116 | 114,115,115.Brewer_Sparrow,train,train,train,train,val,val 117 | 115,116,116.Chipping_Sparrow,train,val,train,val,train,train 118 | 116,117,117.Clay_colored_Sparrow,train,val,train,val,train,train 119 | 117,118,118.House_Sparrow,train,train,val,val,train,train 120 | 118,119,119.Field_Sparrow,test,test,test,test,test,test 121 | 119,120,120.Fox_Sparrow,train,train,train,train,train,train 122 | 120,121,121.Grasshopper_Sparrow,test,test,test,val,val,train 123 | 121,122,122.Harris_Sparrow,val,val,train,train,train,train 124 | 122,123,123.Henslow_Sparrow,train,train,val,test,test,test 125 | 123,124,124.Le_Conte_Sparrow,val,val,train,test,test,test 126 | 124,125,125.Lincoln_Sparrow,val,train,train,train,train,val 127 | 125,126,126.Nelson_Sharp_tailed_Sparrow,train,train,val,train,train,val 128 | 126,127,127.Savannah_Sparrow,train,train,train,test,test,test 129 | 127,128,128.Seaside_Sparrow,train,train,train,val,val,train 130 | 128,129,129.Song_Sparrow,train,train,val,train,train,val 131 | 129,130,130.Tree_Sparrow,test,test,test,test,test,test 132 | 130,131,131.Vesper_Sparrow,val,train,val,train,train,val 133 | 131,132,132.White_crowned_Sparrow,train,val,train,test,test,test 134 | 132,133,133.White_throated_Sparrow,val,val,train,train,val,train 135 | 133,134,134.Cape_Glossy_Starling,val,train,train,train,val,train 136 | 134,135,135.Bank_Swallow,test,test,test,train,val,train 137 | 135,136,136.Barn_Swallow,train,train,train,test,test,test 138 | 136,137,137.Cliff_Swallow,val,val,train,train,train,train 139 | 137,138,138.Tree_Swallow,test,test,test,test,test,test 140 | 138,139,139.Scarlet_Tanager,train,train,train,test,test,test 141 | 139,140,140.Summer_Tanager,val,train,val,val,val,train 142 | 140,141,141.Artic_Tern,train,train,train,val,train,train 143 | 141,142,142.Black_Tern,train,train,val,train,train,val 144 | 142,143,143.Caspian_Tern,val,train,val,test,test,test 145 | 143,144,144.Common_Tern,train,train,train,val,val,val 146 | 144,145,145.Elegant_Tern,val,train,train,train,train,train 147 | 145,146,146.Forsters_Tern,train,train,train,train,train,train 148 | 146,147,147.Least_Tern,test,test,test,train,train,val 149 | 147,148,148.Green_tailed_Towhee,val,train,train,test,test,test 150 | 148,149,149.Brown_Thrasher,train,val,val,train,train,train 151 | 149,150,150.Sage_Thrasher,train,train,train,val,val,train 152 | 150,151,151.Black_capped_Vireo,train,val,val,val,train,train 153 | 151,152,152.Blue_headed_Vireo,val,train,train,train,train,train 154 | 152,153,153.Philadelphia_Vireo,train,val,train,val,train,val 155 | 153,154,154.Red_eyed_Vireo,train,train,train,train,train,train 156 | 154,155,155.Warbling_Vireo,val,val,val,val,train,val 157 | 155,156,156.White_eyed_Vireo,test,test,test,test,test,test 158 | 156,157,157.Yellow_throated_Vireo,val,train,train,test,test,test 159 | 157,158,158.Bay_breasted_Warbler,train,train,train,val,val,train 160 | 158,159,159.Black_and_white_Warbler,train,train,train,train,train,train 161 | 159,160,160.Black_throated_Blue_Warbler,train,train,train,train,train,val 162 | 160,161,161.Blue_winged_Warbler,train,train,train,test,test,test 163 | 161,162,162.Canada_Warbler,train,val,train,val,val,train 164 | 162,163,163.Cape_May_Warbler,test,test,test,test,test,test 165 | 163,164,164.Cerulean_Warbler,val,val,train,test,test,test 166 | 164,165,165.Chestnut_sided_Warbler,test,test,test,test,test,test 167 | 165,166,166.Golden_winged_Warbler,test,test,test,train,train,val 168 | 166,167,167.Hooded_Warbler,train,val,train,train,train,train 169 | 167,168,168.Kentucky_Warbler,train,train,train,test,test,test 170 | 168,169,169.Magnolia_Warbler,val,train,train,test,test,test 171 | 169,170,170.Mourning_Warbler,val,val,val,train,train,train 172 | 170,171,171.Myrtle_Warbler,train,train,train,train,train,val 173 | 171,172,172.Nashville_Warbler,train,train,train,train,val,train 174 | 172,173,173.Orange_crowned_Warbler,train,val,train,test,test,test 175 | 173,174,174.Palm_Warbler,train,train,val,val,val,train 176 | 174,175,175.Pine_Warbler,train,train,val,train,val,val 177 | 175,176,176.Prairie_Warbler,train,train,train,train,train,train 178 | 176,177,177.Prothonotary_Warbler,val,train,train,val,train,train 179 | 177,178,178.Swainson_Warbler,train,train,val,train,val,train 180 | 178,179,179.Tennessee_Warbler,val,train,val,val,val,val 181 | 179,180,180.Wilson_Warbler,test,test,test,test,test,test 182 | 180,181,181.Worm_eating_Warbler,train,val,val,train,val,val 183 | 181,182,182.Yellow_Warbler,train,train,val,val,train,train 184 | 182,183,183.Northern_Waterthrush,test,test,test,train,train,train 185 | 183,184,184.Louisiana_Waterthrush,train,train,train,val,train,val 186 | 184,185,185.Bohemian_Waxwing,test,test,test,train,train,train 187 | 185,186,186.Cedar_Waxwing,test,test,test,train,train,train 188 | 186,187,187.American_Three_toed_Woodpecker,test,test,test,train,val,val 189 | 187,188,188.Pileated_Woodpecker,train,val,train,test,test,test 190 | 188,189,189.Red_bellied_Woodpecker,train,train,val,val,train,train 191 | 189,190,190.Red_cockaded_Woodpecker,train,train,train,test,test,test 192 | 190,191,191.Red_headed_Woodpecker,train,train,val,test,test,test 193 | 191,192,192.Downy_Woodpecker,train,val,train,val,train,train 194 | 192,193,193.Bewick_Wren,train,train,val,train,train,train 195 | 193,194,194.Cactus_Wren,val,train,val,val,train,train 196 | 194,195,195.Carolina_Wren,val,val,val,val,train,train 197 | 195,196,196.House_Wren,train,train,val,train,val,train 198 | 196,197,197.Marsh_Wren,test,test,test,train,val,train 199 | 197,198,198.Rock_Wren,train,train,train,val,train,train 200 | 198,199,199.Winter_Wren,val,train,train,train,train,train 201 | 199,200,200.Common_Yellowthroat,train,train,train,test,test,test 202 | -------------------------------------------------------------------------------- /data/features_resnet101_10crop.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/yannick-lc/iccv2019-triplet-loss/adac3ab2e9b79f7aba1bb170194d9bae39099bcd/data/features_resnet101_10crop.npy -------------------------------------------------------------------------------- /dataset.py: -------------------------------------------------------------------------------- 1 | import pandas as pd 2 | import numpy as np 3 | 4 | class ZslDataset: 5 | 6 | PATHS = { 7 | "awa2": '/scratch_ssd/yannick/Animals_with_Attributes2/preprocessed/', 8 | "cub": '/scratch_ssd/yannick/CUB_200_2011/preprocessed/', 9 | "apy": '/scratch_ssd/yannick/aPascalYahoo/preprocessed/', 10 | "sun": '/scratch_ssd/yannick/SUN/preprocessed/' 11 | } 12 | 13 | def __init__(self, df_classes, df, attributes, features): 14 | """ 15 | N: number of samples in dataset 16 | C: number of classes 17 | D: dimension of visual features space 18 | K: dimension of semantic features space 19 | 20 | Parameters: 21 | df_classes: C-row pandas dataframe representing classes 22 | should contain columns 'class', 'class_name', and splits 23 | df: N-row pandas dataframe representing samples 24 | should contain columns 'class' and splits 25 | attributes: C x K numpy array containing semantic representation of class prototypes 26 | features: N x D numpy array containing visual features representation of samples 27 | """ 28 | self.df_classes = df_classes 29 | self.df = df 30 | self.attributes = attributes 31 | self.features = features -------------------------------------------------------------------------------- /flexible_triplet_loss.py: -------------------------------------------------------------------------------- 1 | """ 2 | Implementation of method in [arxiv link] 3 | 4 | Notations: 5 | 6 | N: number of samples (may be different between train and test) 7 | C: number of classes 8 | D: dimension of visual space 9 | K: dimension of semantic space 10 | H: dimension of learned representation (projected X and S) 11 | 12 | X: usually N x D matrix of visual samples 13 | S: usually C x K matrix of semantic prototypes 14 | """ 15 | 16 | import torch 17 | import torch.nn as nn 18 | import torch.nn.functional as F 19 | import torch.optim as optim 20 | from torch.autograd import Variable 21 | 22 | import sklearn 23 | import numpy as np 24 | import pandas as pd 25 | 26 | 27 | # FLEXIBLE SEMANTIC MARGIN (Section 3.2) 28 | 29 | def pairwise_mahalanobis(S1, S2, Cov_1=None): 30 | """ 31 | S1: C1 x K matrix (torch.FloatTensor) 32 | -> C1 K-dimensional semantic prototypes 33 | S2: C2 x K matrix (torch.FloatTensor) 34 | -> C2 K-dimensional semantic prototypes 35 | Sigma_1: K x K matrix (torch.FloatTensor) 36 | -> inverse of the covariance matrix Sigma; used to compute Mahalanobis distances 37 | by default Sigma is the identity matrix (and so distances are euclidean distances) 38 | 39 | returns an C1 x C2 matrix corresponding to the Mahalanobis distance between each element of S1 and S2 40 | (Equation 5) 41 | """ 42 | if S1.dim() != 2 or S2.dim() != 2 or S1.shape[1] != S2.shape[1]: 43 | raise RuntimeError("Bad input dimension") 44 | C1, K = S1.shape 45 | C2, K = S2.shape 46 | if Cov_1 is None: 47 | Cov_1 = torch.eye(K) 48 | if Cov_1.shape != (K, K): 49 | raise RuntimeError("Bad input dimension") 50 | 51 | S1S2t = S1.matmul(Cov_1).matmul(S2.t()) 52 | S1S1 = S1.matmul(Cov_1).mul(S1).sum(dim=1, keepdim=True).expand(-1, C2) 53 | S2S2 = S2.matmul(Cov_1).mul(S2).sum(dim=1, keepdim=True).t().expand(C1, -1) 54 | return torch.sqrt(torch.abs(S1S1 + S2S2 - 2. * S1S2t) + 1e-32) # to avoid numerical instabilities 55 | 56 | def distance_matrix(S, mahalanobis=True, mean=1., std=0.5): 57 | """ 58 | S: C x K matrix (numpy array) 59 | -> K-dimensional prototypes of C classes 60 | mahalanobis: indicates whether to use Mahalanobis distance (uses euclidean distance if False) 61 | mean & std: target mean and standard deviation 62 | 63 | returns a C x C matrix corresponding to the Mahalanobis distance between each pair of elements of S 64 | rescaled to have approximately target mean and standard deviation while keeping values positive 65 | (Equation 6) 66 | """ 67 | Cov_1 = None 68 | if mahalanobis: 69 | Cov, _ = sklearn.covariance.ledoit_wolf(S) # robust estimation of covariance matrix 70 | Cov_1 = torch.FloatTensor(np.linalg.inv(Cov)) 71 | S = torch.FloatTensor(S) 72 | 73 | distances = pairwise_mahalanobis(S, S, Cov_1) 74 | 75 | # Rescaling to have approximately target mean and standard deviation while keeping values positive 76 | max_zero_distance = distances.diag().max() 77 | positive_distances = np.array([x for x in distances.view(-1) if x > max_zero_distance]) 78 | emp_std = float(positive_distances.std()) 79 | emp_mean = float(positive_distances.mean()) 80 | distances = F.relu(std * (distances - emp_mean) / emp_std + mean) 81 | emp_std = float(distances.std()) 82 | emp_mean = float(distances.mean()) 83 | distances = F.relu(std * (distances - emp_mean) / emp_std + mean) 84 | return distances 85 | 86 | 87 | # PARTIAL NORMALIZATION (Section 3.3) 88 | 89 | def partial_normalization(X, gamma): 90 | """ 91 | X: N x H matrix (torch.FloatTensor) 92 | -> projected visual (or semantic) samples 93 | gamma: scalar between 0 and 1 94 | -> normalization coefficient 95 | 96 | returns N x H matrix corresponding to X matrix where each row has been partially normalize 97 | (Equation 8) 98 | """ 99 | partial_norms = 1. / (gamma * (X.norm(p=2, dim=1) - 1) + 1) 100 | partial_norms = partial_norms.view(-1, 1) 101 | X = partial_norms * X 102 | return X 103 | 104 | 105 | # RELEVANCE WEIGHTING (Section 3.4) 106 | 107 | def class_weights(X_c): 108 | """ 109 | X_c: N x D matrix of N D-dimensional visual samples, assumed to belong to the same class c 110 | 111 | returns the corresponding relevance weights 112 | """ 113 | mean_vector = X_c.mean(axis=0).reshape(-1, 1) 114 | distances_to_mean_vector = np.sqrt((X_c.T - mean_vector).T.dot(X_c.T - mean_vector).diagonal()) 115 | distribution = stats.norm(*scipy.stats.norm.fit(distances_to_mean_vector)) 116 | return 1. - distribution.cdf(distances_to_mean_vector) 117 | 118 | def relevance_weigths(X, Y): 119 | """ 120 | X: N x D matrix of N D-dimensional visual samples 121 | Y: N dimensional vector of classes 122 | 123 | returns an N-dimensional vector corresponding to relevance weights of each visual samples 124 | """ 125 | weights = np.zeros(Y.shape[0]) 126 | classes = sorted(set(Y)) 127 | for c in classes: 128 | indexes_c = np.where(Y == c) 129 | X_c = X[indexes_c] 130 | weigths_c = class_weights(X_c) 131 | weights[indexes_c] = weigths_c 132 | return weights 133 | 134 | 135 | # FINAL MODEL (Section 3.5) 136 | 137 | def flexible_triplet_loss(X_theta, S_psi, Y, V, D_tilde): 138 | """ 139 | X_theta: N x H matrix (torch.FloatTensor) 140 | -> projected visual features 141 | S_psi: C x H matrix (torch.FloatTensor) 142 | -> projected semantic features 143 | Y: N x C binary matrix (torch.LongTensor) 144 | -> labels 145 | V: N-dimensional vector (torch.FloatTensor) 146 | -> relevance weights 147 | Dtilde: C x C (torch.FloatTensor) 148 | -> semantic distance between each class 149 | 150 | returns the corresponding triplet loss 151 | (Equation 13, without regularization omega) 152 | """ 153 | N, H = X_theta.size() 154 | C, _ = S_psi.size() 155 | if DEVICE == "cpu": 156 | Y = Y.type(torch.FloatTensor) 157 | else: 158 | Y = Y.type(torch.cuda.FloatTensor) 159 | 160 | pairwise_compatibilities = X_theta.mm(S_psi.t()) # all the f(x_n, s_c) (Equation 3) 161 | prototype_compatibilities = (Y * pairwise_compatibilities).sum(dim=1).view(-1, 1).expand(-1, C) # all the f(x_n, s_y) (Equation 3) 162 | margin = D_tilde.unsqueeze(0).expand(N, -1, -1) * Y.unsqueeze(2).expand(-1, -1, C) 163 | margin = margin.sum(dim=1) # flexible semantic margin 164 | 165 | triplet_losses = F.relu(margin + pairwise_compatibilities - prototype_compatibilities) # (Equation 12) 166 | triplet_losses = (1. - Y) * triplet_losses # keeping only c != yn (in Equation 13) 167 | triplet_losses = V.view(-1, 1) * triplet_losses # weighting by relevance 168 | loss = triplet_losses.sum() / (N * C) 169 | return loss 170 | 171 | class Projection(nn.Module): 172 | """ 173 | Represents a linear projection from one space (visual or semantic) to another (semantic or common space) 174 | Projections are partially normalized (cf. Section 3.3) 175 | """ 176 | 177 | def __init__(self, d_input, d_embedding, gamma=1.0): 178 | super(Projection, self).__init__() 179 | self.gamma = gamma 180 | self.fc1 = nn.Linear(d_input, d_embedding, bias=True) 181 | 182 | def norm(self): 183 | """ 184 | returns average norm of parameters 185 | """ 186 | norms = ( 187 | self.fc1.weight.norm(p=1) + self.fc1.bias.norm(p=1) 188 | ) 189 | size = ( 190 | self.fc1.weight.shape[0] * self.fc1.weight.shape[1] + self.fc1.bias.shape[0] 191 | ) 192 | return norms / size 193 | 194 | def forward(self, x): 195 | x = self.fc1(x) 196 | x = partial_normalization(x, self.gamma) 197 | return x 198 | 199 | class FlexibleTripletLoss(AbstractModel): 200 | """ 201 | Represents the final model 202 | """ 203 | 204 | def __init__(self, params=None): 205 | # Default parameters 206 | self.params = { 207 | # Hyperparameters 208 | "lambda": 0., # regularization, Equation 13 209 | "mu_dtilde": 1.0, # mean of flexible margin, Equation 6 210 | "sigma_dtilde": 0.5, # standard deviation of flexible margin, Equation 6 211 | "gamma": 1.0, # partial normalization, Equation 8 212 | "setting": "thetapsi", # mapping of visual features (and semantic prototypes), Section 3.5 213 | 214 | # Other options (not hyperparameters) 215 | "epochs": 50, 216 | "learning_rate": 1e-3, 217 | "batch_size": 1000, 218 | "optimizer": optim.Adam, 219 | "num_workers": 4, 220 | "seed": 42L, 221 | "loss_multiplier": 1e3, 222 | "verbose": False, 223 | 224 | # Ablation study 225 | "mahalanobis_distance": True, 226 | "relevance_weighting": True 227 | # to disable partial normalization, set gamma to 0 228 | # to disable flexible semantic margin, set sigma_dtilde to 0 (and possible mu_dtilde to 1) 229 | } 230 | 231 | # Overriding default parameters if specified 232 | if params is not None: 233 | for param in params: 234 | self.params[param] = params[param] 235 | 236 | def fit(self, X, Y, S): 237 | """ 238 | X: N x D matrix (numpy array) 239 | -> visual training features 240 | Y: N-dimensional vector (numpy array) 241 | -> labels 242 | S: C x K matrix (numpy array) 243 | -> training prototypes 244 | """ 245 | torch.manual_seed(self.params["seed"]) 246 | np.random.seed(self.params["seed"]) 247 | 248 | # Relevance weighting, Section 3.4 249 | if self.params["relevance_weighting"]: 250 | V = torch.FloatTensor(relevance_weigths(X, Y)) 251 | else: 252 | V = torch.ones(X.shape[0]) 253 | 254 | N, D = X.shape 255 | C, K = S.shape 256 | H = K # embedding dimension is the dimension of the semantic space, Section 3.5 257 | 258 | X = torch.FloatTensor(X) 259 | Y_ = np.zeros([N, C]) 260 | for n, c in enumerate(Y): 261 | Y_[n, c] = 1 262 | Y = torch.LongTensor(Y_) 263 | self.S = Variable(torch.FloatTensor(S).to(DEVICE)) 264 | 265 | # Flexible semantic margin, Section 3.3 266 | self.D_tilde = Variable(distance_matrix( 267 | S, 268 | mahalanobis=self.params["mahalanobis_distance"], 269 | mean=self.params["mu_dtilde"], std=self.params["mu_dtilde"]*self.params["sigma_dtilde"] 270 | ).to(DEVICE)) 271 | 272 | dataset = torch.utils.data.TensorDataset(X, Y, V) 273 | self.loader = torch.utils.data.DataLoader( 274 | dataset, 275 | batch_size=self.params["batch_size"], shuffle=True, num_workers=self.params["num_workers"] 276 | ) 277 | 278 | # Setting (theta or theta + psi, Section 3.5) 279 | self.visual_projection = Projection(D, H, gamma=self.params["gamma"]).to(DEVICE) 280 | self.semantic_projection = Projection(K, H, gamma=1.0).to(DEVICE) # we always normalize projection of S 281 | 282 | if self.params["setting"] == "thetapsi": # theta + psi 283 | self.optimizer = self.params["optimizer"]( 284 | params=list(self.visual_projection.parameters()) + list(self.semantic_projection.parameters()), 285 | lr=self.params["learning_rate"], weight_decay=0. 286 | ) 287 | else: # only theta 288 | self.optimizer = self.params["optimizer"]( 289 | params=list(self.visual_projection.parameters()), 290 | lr=self.params["learning_rate"], weight_decay=0. 291 | ) 292 | 293 | self.__train() 294 | 295 | def __train(self): 296 | """ 297 | trains the model for the specified number of epochs with the specified hyperparameters (and options) 298 | """ 299 | for epoch in range(self.params["epochs"]): 300 | if self.params["verbose"]: 301 | print "EPOCH %i" % epoch 302 | for i, (inputs, labels, weights) in enumerate(self.loader): 303 | 304 | N, D = inputs.shape 305 | _, C = labels.shape 306 | 307 | X = Variable(inputs.to(DEVICE)) 308 | Y = Variable(labels.to(DEVICE)) 309 | V = Variable(weights.to(DEVICE)) 310 | 311 | self.optimizer.zero_grad() 312 | 313 | X_theta = self.visual_projection(X) 314 | regularization_loss = self.visual_projection.norm() 315 | 316 | if self.params["setting"] == "thetapsi": 317 | S_psi = self.semantic_projection(self.S) 318 | regularization_loss = regularization_loss + self.semantic_projection.norm() 319 | else: 320 | S_psi = self.S 321 | 322 | loss = ( 323 | self.params["loss_multiplier"] * flexible_triplet_loss(X_theta, S_psi, Y, V, self.D_tilde) 324 | + self.params["lambda"] * regularization_loss 325 | ) # Equation 13 326 | loss.backward() 327 | self.optimizer.step() 328 | 329 | if self.params["verbose"]: 330 | print "Loss: %.2f" % loss.item() 331 | 332 | def predict(self, X, S): 333 | """ 334 | X: N x D matrix (numpy array) 335 | -> visual test features 336 | S: C x K matrix (numpy array) 337 | -> test prototypes 338 | 339 | Note: N (number of features) and C (number of classes) are typically not the same as in fit 340 | """ 341 | X = Variable(torch.FloatTensor(X).to(DEVICE)) 342 | S = Variable(torch.FloatTensor(S).to(DEVICE)) 343 | 344 | X_theta = self.visual_projection(X) 345 | if self.params["setting"] == "thetapsi": 346 | S_psi = self.semantic_projection(S) 347 | else: 348 | S_psi = S 349 | 350 | probabilities = X_theta.mm(S_psi.t()) 351 | return probabilities.data.cpu().numpy() 352 | --------------------------------------------------------------------------------