├── .gitignore ├── LICENSE ├── README.md ├── face_detection ├── mediapipe_models │ ├── face_detection_front.tflite │ ├── face_detection_front_4channels.pb │ ├── face_detection_front_4channels.tflite │ ├── face_detection_front_4channels_f16.tflite │ └── face_detection_front_labelmap.txt ├── model.py └── utils.py ├── hair_segmentation ├── mediapipe_models │ └── hair_segmentation.tflite └── model.py ├── hand_landmark ├── assets │ └── hand_21_landmark_index.jpeg ├── mediapipe_models │ ├── hand_landmark.tflite │ ├── hand_landmark_3d.tflite │ ├── hand_landmark_3d_4channels.pb │ ├── hand_landmark_3d_4channels.tflite │ ├── hand_landmark_4channels.pb │ ├── hand_landmark_4channels.tflite │ └── readme.md ├── model_2d.py ├── model_3d.py └── utils.py └── palm_detection ├── assets └── palm_7_landmark_index.png ├── mediapipe_models ├── palm_detection.tflite ├── palm_detection_4channels.pb ├── palm_detection_4channels.tflite └── palm_detection_without_custom_layer.tflite ├── model.py ├── pretrained_weights ├── classificator_16_Bias.npy ├── classificator_16_Kernel.npy ├── classificator_32_Bias.npy ├── classificator_32_Kernel.npy ├── classificator_8_Bias.npy ├── classificator_8_Kernel.npy ├── conv2d_10_Bias.npy ├── conv2d_10_Kernel.npy ├── conv2d_11_Bias.npy ├── conv2d_11_Kernel.npy ├── conv2d_12_Bias.npy ├── conv2d_12_Kernel.npy ├── conv2d_13_Bias.npy ├── conv2d_13_Kernel.npy ├── conv2d_14_Bias.npy ├── conv2d_14_Kernel.npy ├── conv2d_15_Bias.npy ├── conv2d_15_Kernel.npy ├── conv2d_16_Bias.npy ├── conv2d_16_Kernel.npy ├── conv2d_17_Bias.npy ├── conv2d_17_Kernel.npy ├── conv2d_18_Bias.npy ├── conv2d_18_Kernel.npy ├── conv2d_19_Bias.npy ├── conv2d_19_Kernel.npy ├── conv2d_1_Bias.npy ├── conv2d_1_Kernel.npy ├── conv2d_20_Bias.npy ├── conv2d_20_Kernel.npy ├── conv2d_21_Bias.npy ├── conv2d_21_Kernel.npy ├── conv2d_22_Bias.npy ├── conv2d_22_Kernel.npy ├── conv2d_23_Bias.npy ├── conv2d_23_Kernel.npy ├── conv2d_24_Bias.npy ├── conv2d_24_Kernel.npy ├── conv2d_25_Bias.npy ├── conv2d_25_Kernel.npy ├── conv2d_26_Bias.npy ├── conv2d_26_Kernel.npy ├── conv2d_27_Bias.npy ├── conv2d_27_Kernel.npy ├── conv2d_28_Bias.npy ├── conv2d_28_Kernel.npy ├── conv2d_29_Bias.npy ├── conv2d_29_Kernel.npy ├── conv2d_2_Bias.npy ├── conv2d_2_Kernel.npy ├── conv2d_30_Bias.npy ├── conv2d_30_Kernel.npy ├── conv2d_31_Bias.npy ├── conv2d_31_Kernel.npy ├── conv2d_32_Bias.npy ├── conv2d_32_Kernel.npy ├── conv2d_33_Bias.npy ├── conv2d_33_Kernel.npy ├── conv2d_34_Bias.npy ├── conv2d_34_Kernel.npy ├── conv2d_35_Bias.npy ├── conv2d_35_Kernel.npy ├── conv2d_36_Bias.npy ├── conv2d_36_Kernel.npy ├── conv2d_37_Bias.npy ├── conv2d_37_Kernel.npy ├── conv2d_38_Bias.npy ├── conv2d_38_Kernel.npy ├── conv2d_39_Bias.npy ├── conv2d_39_Kernel.npy ├── conv2d_3_Bias.npy ├── conv2d_3_Kernel.npy ├── conv2d_40_Bias.npy ├── conv2d_40_Kernel.npy ├── conv2d_41_Bias.npy ├── conv2d_41_Kernel.npy ├── conv2d_4_Bias.npy ├── conv2d_4_Kernel.npy ├── conv2d_5_Bias.npy ├── conv2d_5_Kernel.npy ├── conv2d_6_Bias.npy ├── conv2d_6_Kernel.npy ├── conv2d_7_Bias.npy ├── conv2d_7_Kernel.npy ├── conv2d_8_Bias.npy ├── conv2d_8_Kernel.npy ├── conv2d_9_Bias.npy ├── conv2d_9_Kernel.npy ├── conv2d_Bias.npy ├── conv2d_Kernel.npy ├── conv2d_transpose_1_Bias.npy ├── conv2d_transpose_1_Kernel.npy ├── conv2d_transpose_Bias.npy ├── conv2d_transpose_Kernel.npy ├── depthwise_conv2d_10_Bias.npy ├── depthwise_conv2d_10_Kernel.npy ├── depthwise_conv2d_11_Bias.npy ├── depthwise_conv2d_11_Kernel.npy ├── depthwise_conv2d_12_Bias.npy ├── depthwise_conv2d_12_Kernel.npy ├── depthwise_conv2d_13_Bias.npy ├── depthwise_conv2d_13_Kernel.npy ├── depthwise_conv2d_14_Bias.npy ├── depthwise_conv2d_14_Kernel.npy ├── depthwise_conv2d_15_Bias.npy ├── depthwise_conv2d_15_Kernel.npy ├── depthwise_conv2d_16_Bias.npy ├── depthwise_conv2d_16_Kernel.npy ├── depthwise_conv2d_17_Bias.npy ├── depthwise_conv2d_17_Kernel.npy ├── depthwise_conv2d_18_Bias.npy ├── depthwise_conv2d_18_Kernel.npy ├── depthwise_conv2d_19_Bias.npy ├── depthwise_conv2d_19_Kernel.npy ├── depthwise_conv2d_1_Bias.npy ├── depthwise_conv2d_1_Kernel.npy ├── depthwise_conv2d_20_Bias.npy ├── depthwise_conv2d_20_Kernel.npy ├── depthwise_conv2d_21_Bias.npy ├── depthwise_conv2d_21_Kernel.npy ├── depthwise_conv2d_22_Bias.npy ├── depthwise_conv2d_22_Kernel.npy ├── depthwise_conv2d_23_Bias.npy ├── depthwise_conv2d_23_Kernel.npy ├── depthwise_conv2d_24_Bias.npy ├── depthwise_conv2d_24_Kernel.npy ├── depthwise_conv2d_25_Bias.npy ├── depthwise_conv2d_25_Kernel.npy ├── depthwise_conv2d_26_Bias.npy ├── depthwise_conv2d_26_Kernel.npy ├── depthwise_conv2d_27_Bias.npy ├── depthwise_conv2d_27_Kernel.npy ├── depthwise_conv2d_28_Bias.npy ├── depthwise_conv2d_28_Kernel.npy ├── depthwise_conv2d_29_Bias.npy ├── depthwise_conv2d_29_Kernel.npy ├── depthwise_conv2d_2_Bias.npy ├── depthwise_conv2d_2_Kernel.npy ├── depthwise_conv2d_30_Bias.npy ├── depthwise_conv2d_30_Kernel.npy ├── depthwise_conv2d_31_Bias.npy ├── depthwise_conv2d_31_Kernel.npy ├── depthwise_conv2d_32_Bias.npy ├── depthwise_conv2d_32_Kernel.npy ├── depthwise_conv2d_33_Bias.npy ├── depthwise_conv2d_33_Kernel.npy ├── depthwise_conv2d_34_Bias.npy ├── depthwise_conv2d_34_Kernel.npy ├── depthwise_conv2d_35_Bias.npy ├── depthwise_conv2d_35_Kernel.npy ├── depthwise_conv2d_36_Bias.npy ├── depthwise_conv2d_36_Kernel.npy ├── depthwise_conv2d_37_Bias.npy ├── depthwise_conv2d_37_Kernel.npy ├── depthwise_conv2d_38_Bias.npy ├── depthwise_conv2d_38_Kernel.npy ├── depthwise_conv2d_39_Bias.npy ├── depthwise_conv2d_39_Kernel.npy ├── depthwise_conv2d_3_Bias.npy ├── depthwise_conv2d_3_Kernel.npy ├── depthwise_conv2d_40_Bias.npy ├── depthwise_conv2d_40_Kernel.npy ├── depthwise_conv2d_4_Bias.npy ├── depthwise_conv2d_4_Kernel.npy ├── depthwise_conv2d_5_Bias.npy ├── depthwise_conv2d_5_Kernel.npy ├── depthwise_conv2d_6_Bias.npy ├── depthwise_conv2d_6_Kernel.npy ├── depthwise_conv2d_7_Bias.npy ├── depthwise_conv2d_7_Kernel.npy ├── depthwise_conv2d_8_Bias.npy ├── depthwise_conv2d_8_Kernel.npy ├── depthwise_conv2d_9_Bias.npy ├── depthwise_conv2d_9_Kernel.npy ├── depthwise_conv2d_Bias.npy ├── depthwise_conv2d_Kernel.npy ├── regressor_16_Bias.npy ├── regressor_16_Kernel.npy ├── regressor_32_Bias.npy ├── regressor_32_Kernel.npy ├── regressor_8_Bias.npy └── regressor_8_Kernel.npy └── utils.py /.gitignore: -------------------------------------------------------------------------------- 1 | # Byte-compiled / optimized / DLL files 2 | __pycache__/ 3 | *.py[cod] 4 | *$py.class 5 | 6 | # C extensions 7 | *.so 8 | 9 | # Distribution / packaging 10 | .Python 11 | build/ 12 | develop-eggs/ 13 | dist/ 14 | downloads/ 15 | eggs/ 16 | .eggs/ 17 | lib/ 18 | lib64/ 19 | parts/ 20 | sdist/ 21 | var/ 22 | wheels/ 23 | pip-wheel-metadata/ 24 | share/python-wheels/ 25 | *.egg-info/ 26 | .installed.cfg 27 | *.egg 28 | MANIFEST 29 | 30 | # PyInstaller 31 | # Usually these files are written by a python script from a template 32 | # before PyInstaller builds the exe, so as to inject date/other infos into it. 33 | *.manifest 34 | *.spec 35 | 36 | # Installer logs 37 | pip-log.txt 38 | pip-delete-this-directory.txt 39 | 40 | # Unit test / coverage reports 41 | htmlcov/ 42 | .tox/ 43 | .nox/ 44 | .coverage 45 | .coverage.* 46 | .cache 47 | nosetests.xml 48 | coverage.xml 49 | *.cover 50 | *.py,cover 51 | .hypothesis/ 52 | .pytest_cache/ 53 | 54 | # Translations 55 | *.mo 56 | *.pot 57 | 58 | # Django stuff: 59 | *.log 60 | local_settings.py 61 | db.sqlite3 62 | db.sqlite3-journal 63 | 64 | # Flask stuff: 65 | instance/ 66 | .webassets-cache 67 | 68 | # Scrapy stuff: 69 | .scrapy 70 | 71 | # Sphinx documentation 72 | docs/_build/ 73 | 74 | # PyBuilder 75 | target/ 76 | 77 | # Jupyter Notebook 78 | .ipynb_checkpoints 79 | 80 | # IPython 81 | profile_default/ 82 | ipython_config.py 83 | 84 | # pyenv 85 | .python-version 86 | 87 | # pipenv 88 | # According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control. 89 | # However, in case of collaboration, if having platform-specific dependencies or dependencies 90 | # having no cross-platform support, pipenv may install dependencies that don't work, or not 91 | # install all needed dependencies. 92 | #Pipfile.lock 93 | 94 | # PEP 582; used by e.g. github.com/David-OConnor/pyflow 95 | __pypackages__/ 96 | 97 | # Celery stuff 98 | celerybeat-schedule 99 | celerybeat.pid 100 | 101 | # SageMath parsed files 102 | *.sage.py 103 | 104 | # Environments 105 | .env 106 | .venv 107 | env/ 108 | venv/ 109 | ENV/ 110 | env.bak/ 111 | venv.bak/ 112 | 113 | # Spyder project settings 114 | .spyderproject 115 | .spyproject 116 | 117 | # Rope project settings 118 | .ropeproject 119 | 120 | # mkdocs documentation 121 | /site 122 | 123 | # mypy 124 | .mypy_cache/ 125 | .dmypy.json 126 | dmypy.json 127 | 128 | # Pyre type checker 129 | .pyre/ 130 | -------------------------------------------------------------------------------- /LICENSE: -------------------------------------------------------------------------------- 1 | Apache License 2 | Version 2.0, January 2004 3 | http://www.apache.org/licenses/ 4 | 5 | TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION 6 | 7 | 1. Definitions. 8 | 9 | "License" shall mean the terms and conditions for use, reproduction, 10 | and distribution as defined by Sections 1 through 9 of this document. 11 | 12 | "Licensor" shall mean the copyright owner or entity authorized by 13 | the copyright owner that is granting the License. 14 | 15 | "Legal Entity" shall mean the union of the acting entity and all 16 | other entities that control, are controlled by, or are under common 17 | control with that entity. For the purposes of this definition, 18 | "control" means (i) the power, direct or indirect, to cause the 19 | direction or management of such entity, whether by contract or 20 | otherwise, or (ii) ownership of fifty percent (50%) or more of the 21 | outstanding shares, or (iii) beneficial ownership of such entity. 22 | 23 | "You" (or "Your") shall mean an individual or Legal Entity 24 | exercising permissions granted by this License. 25 | 26 | "Source" form shall mean the preferred form for making modifications, 27 | including but not limited to software source code, documentation 28 | source, and configuration files. 29 | 30 | "Object" form shall mean any form resulting from mechanical 31 | transformation or translation of a Source form, including but 32 | not limited to compiled object code, generated documentation, 33 | and conversions to other media types. 34 | 35 | "Work" shall mean the work of authorship, whether in Source or 36 | Object form, made available under the License, as indicated by a 37 | copyright notice that is included in or attached to the work 38 | (an example is provided in the Appendix below). 39 | 40 | "Derivative Works" shall mean any work, whether in Source or Object 41 | form, that is based on (or derived from) the Work and for which the 42 | editorial revisions, annotations, elaborations, or other modifications 43 | represent, as a whole, an original work of authorship. For the purposes 44 | of this License, Derivative Works shall not include works that remain 45 | separable from, or merely link (or bind by name) to the interfaces of, 46 | the Work and Derivative Works thereof. 47 | 48 | "Contribution" shall mean any work of authorship, including 49 | the original version of the Work and any modifications or additions 50 | to that Work or Derivative Works thereof, that is intentionally 51 | submitted to Licensor for inclusion in the Work by the copyright owner 52 | or by an individual or Legal Entity authorized to submit on behalf of 53 | the copyright owner. For the purposes of this definition, "submitted" 54 | means any form of electronic, verbal, or written communication sent 55 | to the Licensor or its representatives, including but not limited to 56 | communication on electronic mailing lists, source code control systems, 57 | and issue tracking systems that are managed by, or on behalf of, the 58 | Licensor for the purpose of discussing and improving the Work, but 59 | excluding communication that is conspicuously marked or otherwise 60 | designated in writing by the copyright owner as "Not a Contribution." 61 | 62 | "Contributor" shall mean Licensor and any individual or Legal Entity 63 | on behalf of whom a Contribution has been received by Licensor and 64 | subsequently incorporated within the Work. 65 | 66 | 2. Grant of Copyright License. Subject to the terms and conditions of 67 | this License, each Contributor hereby grants to You a perpetual, 68 | worldwide, non-exclusive, no-charge, royalty-free, irrevocable 69 | copyright license to reproduce, prepare Derivative Works of, 70 | publicly display, publicly perform, sublicense, and distribute the 71 | Work and such Derivative Works in Source or Object form. 72 | 73 | 3. Grant of Patent License. Subject to the terms and conditions of 74 | this License, each Contributor hereby grants to You a perpetual, 75 | worldwide, non-exclusive, no-charge, royalty-free, irrevocable 76 | (except as stated in this section) patent license to make, have made, 77 | use, offer to sell, sell, import, and otherwise transfer the Work, 78 | where such license applies only to those patent claims licensable 79 | by such Contributor that are necessarily infringed by their 80 | Contribution(s) alone or by combination of their Contribution(s) 81 | with the Work to which such Contribution(s) was submitted. If You 82 | institute patent litigation against any entity (including a 83 | cross-claim or counterclaim in a lawsuit) alleging that the Work 84 | or a Contribution incorporated within the Work constitutes direct 85 | or contributory patent infringement, then any patent licenses 86 | granted to You under this License for that Work shall terminate 87 | as of the date such litigation is filed. 88 | 89 | 4. Redistribution. You may reproduce and distribute copies of the 90 | Work or Derivative Works thereof in any medium, with or without 91 | modifications, and in Source or Object form, provided that You 92 | meet the following conditions: 93 | 94 | (a) You must give any other recipients of the Work or 95 | Derivative Works a copy of this License; and 96 | 97 | (b) You must cause any modified files to carry prominent notices 98 | stating that You changed the files; and 99 | 100 | (c) You must retain, in the Source form of any Derivative Works 101 | that You distribute, all copyright, patent, trademark, and 102 | attribution notices from the Source form of the Work, 103 | excluding those notices that do not pertain to any part of 104 | the Derivative Works; and 105 | 106 | (d) If the Work includes a "NOTICE" text file as part of its 107 | distribution, then any Derivative Works that You distribute must 108 | include a readable copy of the attribution notices contained 109 | within such NOTICE file, excluding those notices that do not 110 | pertain to any part of the Derivative Works, in at least one 111 | of the following places: within a NOTICE text file distributed 112 | as part of the Derivative Works; within the Source form or 113 | documentation, if provided along with the Derivative Works; or, 114 | within a display generated by the Derivative Works, if and 115 | wherever such third-party notices normally appear. The contents 116 | of the NOTICE file are for informational purposes only and 117 | do not modify the License. You may add Your own attribution 118 | notices within Derivative Works that You distribute, alongside 119 | or as an addendum to the NOTICE text from the Work, provided 120 | that such additional attribution notices cannot be construed 121 | as modifying the License. 122 | 123 | You may add Your own copyright statement to Your modifications and 124 | may provide additional or different license terms and conditions 125 | for use, reproduction, or distribution of Your modifications, or 126 | for any such Derivative Works as a whole, provided Your use, 127 | reproduction, and distribution of the Work otherwise complies with 128 | the conditions stated in this License. 129 | 130 | 5. Submission of Contributions. Unless You explicitly state otherwise, 131 | any Contribution intentionally submitted for inclusion in the Work 132 | by You to the Licensor shall be under the terms and conditions of 133 | this License, without any additional terms or conditions. 134 | Notwithstanding the above, nothing herein shall supersede or modify 135 | the terms of any separate license agreement you may have executed 136 | with Licensor regarding such Contributions. 137 | 138 | 6. Trademarks. This License does not grant permission to use the trade 139 | names, trademarks, service marks, or product names of the Licensor, 140 | except as required for reasonable and customary use in describing the 141 | origin of the Work and reproducing the content of the NOTICE file. 142 | 143 | 7. Disclaimer of Warranty. Unless required by applicable law or 144 | agreed to in writing, Licensor provides the Work (and each 145 | Contributor provides its Contributions) on an "AS IS" BASIS, 146 | WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or 147 | implied, including, without limitation, any warranties or conditions 148 | of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A 149 | PARTICULAR PURPOSE. You are solely responsible for determining the 150 | appropriateness of using or redistributing the Work and assume any 151 | risks associated with Your exercise of permissions under this License. 152 | 153 | 8. Limitation of Liability. In no event and under no legal theory, 154 | whether in tort (including negligence), contract, or otherwise, 155 | unless required by applicable law (such as deliberate and grossly 156 | negligent acts) or agreed to in writing, shall any Contributor be 157 | liable to You for damages, including any direct, indirect, special, 158 | incidental, or consequential damages of any character arising as a 159 | result of this License or out of the use or inability to use the 160 | Work (including but not limited to damages for loss of goodwill, 161 | work stoppage, computer failure or malfunction, or any and all 162 | other commercial damages or losses), even if such Contributor 163 | has been advised of the possibility of such damages. 164 | 165 | 9. Accepting Warranty or Additional Liability. While redistributing 166 | the Work or Derivative Works thereof, You may choose to offer, 167 | and charge a fee for, acceptance of support, warranty, indemnity, 168 | or other liability obligations and/or rights consistent with this 169 | License. However, in accepting such obligations, You may act only 170 | on Your own behalf and on Your sole responsibility, not on behalf 171 | of any other Contributor, and only if You agree to indemnify, 172 | defend, and hold each Contributor harmless for any liability 173 | incurred by, or claims asserted against, such Contributor by reason 174 | of your accepting any such warranty or additional liability. 175 | 176 | END OF TERMS AND CONDITIONS 177 | 178 | APPENDIX: How to apply the Apache License to your work. 179 | 180 | To apply the Apache License to your work, attach the following 181 | boilerplate notice, with the fields enclosed by brackets "[]" 182 | replaced with your own identifying information. (Don't include 183 | the brackets!) The text should be enclosed in the appropriate 184 | comment syntax for the file format. We also recommend that a 185 | file or class name and description of purpose be included on the 186 | same "printed page" as the copyright notice for easier 187 | identification within third-party archives. 188 | 189 | Copyright [yyyy] [name of copyright owner] 190 | 191 | Licensed under the Apache License, Version 2.0 (the "License"); 192 | you may not use this file except in compliance with the License. 193 | You may obtain a copy of the License at 194 | 195 | http://www.apache.org/licenses/LICENSE-2.0 196 | 197 | Unless required by applicable law or agreed to in writing, software 198 | distributed under the License is distributed on an "AS IS" BASIS, 199 | WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 200 | See the License for the specific language governing permissions and 201 | limitations under the License. 202 | -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # mediapipe-models 2 | Google Mediapipe public TFLite models implemented using Tensorflow-keras (https://mediapipe.dev) 3 | 4 | ### Requirements 5 | - Tensorflow - 1.13.1 6 | 7 | ### Implementation process 8 | 1. Hand Tracking 9 | * Related links 10 | - [Google AI blog post: On device real time hand tracking](https://ai.googleblog.com/2019/08/on-device-real-time-hand-tracking-with.html) 11 | - Github: [Mediapipe: Hand tracking](https://github.com/google/mediapipe/blob/master/mediapipe/docs/hand_tracking_mobile_gpu.md) 12 | 13 | * Hand Landmark 2D,3D 14 | - Model Architecture (OK) 15 | - Set Pretrained weights (OK) 16 | - Convert TFLite model for 4 channels input (OK) 17 | 18 | * Palm(Hand) Detection 19 | - Model Architecture (OK) 20 | - Set Pretrained weights (OK) 21 | - Convert TFLite model for 4 channels input (OK) 22 | 23 | 2. Face Detection 24 | * Related links 25 | - Paper: ["BlazeFace: Sub-millisecond Neural Face Detection on Mobile GPUs"](https://arxiv.org/abs/1907.05047) 26 | - Github: [Mediapipe: Face detection](https://github.com/google/mediapipe/blob/master/mediapipe/docs/face_detection_mobile_gpu.md) 27 | 28 | * Face Detection 29 | - Model Architecture (OK) 30 | - Set Pretrained weights (TODO) 31 | - Convert TFLite model for 4 channels input (TODO) 32 | 33 | 3. Hair Segmentation 34 | * Related links 35 | - Paper: ["Real-time Hair segmentation and recoloring on Mobile GPUs"](https://arxiv.org/abs/1907.06740) 36 | - Github: [Mediapipe: Hair segmentation](https://github.com/google/mediapipe/blob/master/mediapipe/docs/hair_segmentation_mobile_gpu.md) 37 | 38 | * Hair Segmentation 39 | - Model Architecture (TODO) 40 | - Set Pretrained weights (TODO) 41 | - Convert TFLite model for 4 channels input (TODO) 42 | -------------------------------------------------------------------------------- /face_detection/mediapipe_models/face_detection_front.tflite: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/face_detection/mediapipe_models/face_detection_front.tflite -------------------------------------------------------------------------------- /face_detection/mediapipe_models/face_detection_front_4channels.pb: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/face_detection/mediapipe_models/face_detection_front_4channels.pb -------------------------------------------------------------------------------- /face_detection/mediapipe_models/face_detection_front_4channels.tflite: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/face_detection/mediapipe_models/face_detection_front_4channels.tflite -------------------------------------------------------------------------------- /face_detection/mediapipe_models/face_detection_front_4channels_f16.tflite: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/face_detection/mediapipe_models/face_detection_front_4channels_f16.tflite -------------------------------------------------------------------------------- /face_detection/mediapipe_models/face_detection_front_labelmap.txt: -------------------------------------------------------------------------------- 1 | Face 2 | -------------------------------------------------------------------------------- /face_detection/model.py: -------------------------------------------------------------------------------- 1 | # BlazeFace: Sub-millisecond Neural Face Detection on Mobile GPUs (https://arxiv.org/abs/1907.05047) 2 | ## but it's "not same architecture" as shown in the above. 3 | 4 | import tensorflow as tf 5 | from tensorflow.keras.models import Model 6 | from tensorflow.keras.layers import Input, Conv2D, Add, ReLU, MaxPooling2D, Reshape, Lambda, Activation, DepthwiseConv2D, Concatenate 7 | import numpy as np 8 | 9 | def conv_blocks(x, num_filter, channel_padding=False, pad_value=None): 10 | x = ReLU()(x) 11 | shortcut = x 12 | 13 | if channel_padding: 14 | paddings = tf.constant([[0,0],[0,0],[0,0],[0,pad_value]]) 15 | shortcut = Lambda(lambda x: tf.pad(x, paddings, "CONSTANT"))(shortcut) 16 | 17 | x = DepthwiseConv2D(kernel_size=(3, 3), strides=(1, 1), padding='same', use_bias=True)(x) 18 | x = Conv2D(num_filter, kernel_size=(1, 1), strides=(1, 1), padding='valid', use_bias=True)(x) 19 | x = Add()([x, shortcut]) 20 | return x 21 | 22 | def conv_blocks_with_pooling(x, num_filter, pad_value=None): 23 | x = ReLU()(x) 24 | shortcut = x 25 | 26 | shortcut = MaxPooling2D(strides=(2, 2), padding='same')(shortcut) 27 | paddings = tf.constant([[0,0],[0,0],[0,0],[0,pad_value]]) 28 | shortcut = Lambda(lambda x: tf.pad(x, paddings, "CONSTANT"))(shortcut) 29 | 30 | x = DepthwiseConv2D(kernel_size=(3, 3), strides=(2, 2), padding='same', use_bias=True)(x) 31 | x = Conv2D(num_filter, kernel_size=(1, 1), strides=(1, 1), padding='valid', use_bias=True)(x) 32 | x = Add()([x, shortcut]) 33 | return x 34 | 35 | def face_detection_model(input_size=(128, 128, 3)): 36 | inputs = Input(input_size) 37 | x = Conv2D(24, kernel_size=(5, 5), strides=(2, 2), padding='same', use_bias=True)(inputs) # (1, 64, 64, 24) 38 | x = conv_blocks(x, 24, channel_padding=False) # (1, 64, 64, 24) 39 | x = conv_blocks(x, 28, channel_padding=True, pad_value=4) # (1, 64, 64, 28) 40 | x = conv_blocks_with_pooling(x, 32, pad_value=4) # (1, 32, 32, 32) 41 | x = conv_blocks(x, 36, channel_padding=True, pad_value=4) # (1, 32, 32, 36) 42 | x = conv_blocks(x, 42, channel_padding=True, pad_value=6) # (1, 32, 32, 42) 43 | x = conv_blocks_with_pooling(x, 48, pad_value=6) # (1, 16, 16, 48) 44 | x = conv_blocks(x, 56, channel_padding=True, pad_value=8) # (1, 16, 16, 56) 45 | x = conv_blocks(x, 64, channel_padding=True, pad_value=8) # (1, 16, 16, 64) 46 | x = conv_blocks(x, 72, channel_padding=True, pad_value=8) # (1, 16, 16, 72) 47 | x = conv_blocks(x, 80, channel_padding=True, pad_value=8) # (1, 16, 16, 80) 48 | x = conv_blocks(x, 88, channel_padding=True, pad_value=8) # (1, 16, 16, 88) 49 | 50 | x = ReLU()(x) 51 | shortcut_1 = x # (1, 16, 16, 88) 52 | shortcut = x 53 | 54 | shortcut = MaxPooling2D(strides=(2, 2), padding='same')(shortcut) 55 | paddings = tf.constant([[0,0],[0,0],[0,0],[0,8]]) # pad_value = 8 56 | shortcut = Lambda(lambda x: tf.pad(x, paddings, "CONSTANT"))(shortcut) 57 | 58 | x = DepthwiseConv2D(kernel_size=(3, 3), strides=(2, 2), padding='same', use_bias=True)(x) 59 | x = Conv2D(96, kernel_size=(1, 1), strides=(1, 1), padding='valid', use_bias=True)(x) 60 | x = Add()([x, shortcut]) # (1, 8, 8, 96) 61 | 62 | x = conv_blocks(x, 96, channel_padding=False) # (1, 8, 8, 96) 63 | x = conv_blocks(x, 96, channel_padding=False) # (1, 8, 8, 96) 64 | x = conv_blocks(x, 96, channel_padding=False) # (1, 8, 8, 96) 65 | x = conv_blocks(x, 96, channel_padding=False) # (1, 8, 8, 96) 66 | 67 | # Last layer 68 | x = ReLU()(x) 69 | ## Classificators 70 | classificator_8 = Conv2D(2, kernel_size=(1, 1), strides=(1, 1), padding='same', use_bias=True, name='classificator_8')(shortcut_1) 71 | classificator_16 = Conv2D(6, kernel_size=(1, 1), strides=(1, 1), padding='same', use_bias=True, name='classificator_16')(x) 72 | ### Classificators: Reshape and Concatenation 73 | classificator_8 = Reshape(target_shape=(-1,1))(classificator_8) # (1, 512, 1) 74 | classificator_16 = Reshape(target_shape=(-1,1))(classificator_16) # (1, 384, 1) 75 | classificator_concat = Concatenate(axis=1, name='classificators')([classificator_8, classificator_16]) 76 | 77 | ## Regressors 78 | regressor_8 = Conv2D(32, kernel_size=(1, 1), strides=(1, 1), padding='same', use_bias=True, name='regressor_8')(shortcut_1) 79 | regressor_16 = Conv2D(96, kernel_size=(1, 1), strides=(1, 1), padding='same', use_bias=True, name='regressor_16')(x) 80 | ### Regressors: Reshape and Concatenation 81 | regressor_8 = Reshape(target_shape=(-1,16))(regressor_8) # (1, 512, 1) 82 | regressor_16 = Reshape(target_shape=(-1,16))(regressor_16) # (1, 384, 1) 83 | regressor_concat = Concatenate(axis=1, name='regressors')([regressor_8, regressor_16]) 84 | 85 | model = Model(inputs, [regressor_concat, classificator_concat]) 86 | return model -------------------------------------------------------------------------------- /face_detection/utils.py: -------------------------------------------------------------------------------- 1 | import tensorflow as tf 2 | import numpy as np 3 | import tensorflow.keras.backend as K 4 | 5 | def define_fake_4_channels_graph(input_size=(128, 128, 4)): 6 | inputs = tf.keras.layers.Input(input_size) 7 | x = tf.keras.layers.Lambda(lambda x: x[:,:,:,:3], name='slicing_inputs')(inputs) # slicing into 3-channels 8 | model = tf.keras.models.Model(inputs, x) 9 | sess = K.get_session() 10 | fake_graph = sess.graph_def 11 | K.clear_session() 12 | return fake_graph 13 | 14 | def get_pretrained_tflite_weights(model_path): 15 | model = tf.lite.Interpreter(model_path) 16 | tensor_details = model.get_tensor_details() 17 | weights_dict = {} 18 | layer_names = [] 19 | 20 | for idx in range(0, len(tensor_details)): 21 | try: 22 | name = tensor_details[idx]['name'] 23 | weights = model.get_tensor(idx) 24 | weights_dict[name] = weights 25 | except: 26 | name = tensor_details[idx]['name'] 27 | layer_names.append(name) 28 | return weights_dict, layer_names 29 | 30 | def display_nodes(nodes): 31 | for i, node in enumerate(nodes): 32 | print('%d %s %s' % (i, node.name, node.op)) 33 | for idx, n in enumerate(node.input): 34 | print(u'└─── %d ─ %s' % (idx, n)) 35 | 36 | def set_pretrained_weights(model, weights_dict, layer_names): 37 | for name in layer_names: 38 | if name.find('conv') != -1: 39 | pretrained_weights = [] 40 | kernel_weight = weights_dict.get(name+'/Kernel') 41 | bias_weight = weights_dict.get(name+'/Bias') 42 | kernel_weight = kernel_weight.transpose(1, 2, 3, 0) 43 | pretrained_weights.append(kernel_weight) 44 | pretrained_weights.append(bias_weight) 45 | layer = model.get_layer(name) 46 | layer.set_weights(pretrained_weights) 47 | 48 | elif name.find('classificator_') != -1: 49 | pretrained_weights = [] 50 | kernel_weight = weights_dict.get(name+'/Kernel') 51 | bias_weight = weights_dict.get(name+'/Bias') 52 | kernel_weight = kernel_weight.transpose(1, 2, 3, 0) 53 | pretrained_weights.append(kernel_weight) 54 | pretrained_weights.append(bias_weight) 55 | layer = model.get_layer(name) 56 | layer.set_weights(pretrained_weights) 57 | 58 | elif name.find('regressor_') != -1: 59 | pretrained_weights = [] 60 | kernel_weight = weights_dict.get(name+'/Kernel') 61 | bias_weight = weights_dict.get(name+'/Bias') 62 | kernel_weight = kernel_weight.transpose(1, 2, 3, 0) 63 | pretrained_weights.append(kernel_weight) 64 | pretrained_weights.append(bias_weight) 65 | layer = model.get_layer(name) 66 | layer.set_weights(pretrained_weights) 67 | print("[INFO] Set all pretrained weights") 68 | 69 | def convert_to_pb(model, out_path): 70 | sess = K.get_session() 71 | output_names = [node.op.name for node in model.outputs] 72 | frozen_def = tf.graph_util.convert_variables_to_constants(sess, sess.graph_def, output_names) 73 | with tf.gfile.GFile(out_path, 'w') as f: 74 | f.write(frozen_def.SerializeToString()) 75 | print("[INFO] Save frozen graph model in %s"%(out_path)) 76 | -------------------------------------------------------------------------------- /hair_segmentation/mediapipe_models/hair_segmentation.tflite: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/hair_segmentation/mediapipe_models/hair_segmentation.tflite -------------------------------------------------------------------------------- /hair_segmentation/model.py: -------------------------------------------------------------------------------- 1 | # Real-time Hair Segmentation and Recoloring on Mobile GPUs (https://arxiv.org/abs/1907.06740) 2 | import tensorflow as tf 3 | # from tensorflow.keras.layers import Input, Conv2D, Conv2DTranspose, PReLU, MaxPooling2D, Add, Concatenate 4 | from keras.layers import Input, Conv2D, Conv2DTranspose, PReLU, MaxPooling2D, Add, Concatenate 5 | ## MaxUnpooling2D --> Input: tensor and Argmaxed Tensor --> How to? 6 | ## MaxPoolingWithArgmax2D --> Compatible with tf.nn.max_pool_with_argmax 7 | ## 8 | ## TODO 9 | -------------------------------------------------------------------------------- /hand_landmark/assets/hand_21_landmark_index.jpeg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/hand_landmark/assets/hand_21_landmark_index.jpeg -------------------------------------------------------------------------------- /hand_landmark/mediapipe_models/hand_landmark.tflite: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/hand_landmark/mediapipe_models/hand_landmark.tflite -------------------------------------------------------------------------------- /hand_landmark/mediapipe_models/hand_landmark_3d.tflite: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/hand_landmark/mediapipe_models/hand_landmark_3d.tflite -------------------------------------------------------------------------------- /hand_landmark/mediapipe_models/hand_landmark_3d_4channels.pb: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/hand_landmark/mediapipe_models/hand_landmark_3d_4channels.pb -------------------------------------------------------------------------------- /hand_landmark/mediapipe_models/hand_landmark_3d_4channels.tflite: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/hand_landmark/mediapipe_models/hand_landmark_3d_4channels.tflite -------------------------------------------------------------------------------- /hand_landmark/mediapipe_models/hand_landmark_4channels.pb: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/hand_landmark/mediapipe_models/hand_landmark_4channels.pb -------------------------------------------------------------------------------- /hand_landmark/mediapipe_models/hand_landmark_4channels.tflite: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/hand_landmark/mediapipe_models/hand_landmark_4channels.tflite -------------------------------------------------------------------------------- /hand_landmark/mediapipe_models/readme.md: -------------------------------------------------------------------------------- 1 | - The Pretrained models can download in https://github.com/google/mediapipe/ 2 | * hand_landmark_3d.tflite 3 | * hand_landmark.tflite -------------------------------------------------------------------------------- /hand_landmark/model_2d.py: -------------------------------------------------------------------------------- 1 | # Hand landmark 2d model 2 | from tensorflow.keras.layers import Input, Conv2D, PReLU, SeparableConv2D, Add, ReLU, MaxPooling2D, Reshape, Lambda, Activation, DepthwiseConv2D 3 | from tensorflow.keras.activations import sigmoid 4 | from tensorflow.keras.models import Model 5 | import tensorflow as tf 6 | 7 | def conv_blocks(x, num_filter, activation='prelu', num_iterations=1): 8 | for num_iter in range(0, num_iterations): 9 | if activation == 'prelu': 10 | x = PReLU(alpha_initializer='uniform', shared_axes=[1, 2])(x) 11 | elif activation == 'relu': 12 | x = ReLU()(x) 13 | 14 | shortcut = x 15 | x = Conv2D(int(num_filter / 2), kernel_size=(1, 1), strides=(1, 1), padding='valid', use_bias=True)(x) 16 | x = PReLU(alpha_initializer='uniform', shared_axes=[1, 2])(x) 17 | x = DepthwiseConv2D(kernel_size=(3, 3), strides=(1, 1), padding='same', use_bias=True)(x) 18 | x = Conv2D(num_filter, kernel_size=(1, 1), strides=(1, 1), padding='valid', use_bias=True)(x) 19 | x = Add()([shortcut, x]) 20 | return x 21 | 22 | def conv_blocks_with_pooling(x, num_filter, activation='prelu', padding=False, pad_value=None): 23 | if activation == 'prelu': 24 | x = PReLU(alpha_initializer='uniform', shared_axes=[1, 2])(x) 25 | elif activation == 'relu': 26 | x = ReLU()(x) 27 | 28 | shortcut = x 29 | shortcut = MaxPooling2D(pool_size=(2, 2), strides=(2, 2), padding='valid')(shortcut) 30 | 31 | if padding: 32 | paddings = tf.constant([[0,0],[0,0],[0,0],[0,pad_value]]) 33 | shortcut = Lambda(lambda x: tf.pad(x, paddings, "CONSTANT"))(shortcut) # CONSTANT / REFLECT / SYMMETRIC 34 | x = Conv2D(int(num_filter / 2), kernel_size=(2, 2), strides=(2, 2), padding='valid', use_bias=True)(x) 35 | x = PReLU(alpha_initializer='uniform', shared_axes=[1, 2])(x) 36 | x = DepthwiseConv2D(kernel_size=(3, 3), strides=(1, 1), padding='same', use_bias=True)(x) 37 | x = Conv2D(num_filter, kernel_size=(1, 1), strides=(1, 1), padding='valid', use_bias=True)(x) 38 | x = Add()([x, shortcut]) 39 | else: 40 | x = Conv2D(int(num_filter / 2), kernel_size=(2, 2), strides=(2, 2), padding='valid', use_bias=True)(x) 41 | x = PReLU(alpha_initializer='uniform', shared_axes=[1, 2])(x) 42 | x = DepthwiseConv2D(kernel_size=(3, 3), strides=(1, 1), padding='same', use_bias=True)(x) 43 | x = Conv2D(num_filter, kernel_size=(1, 1), strides=(1, 1), padding='valid', use_bias=True)(x) 44 | x = Add()([x, shortcut]) 45 | return x 46 | 47 | def hand_landmark_2d_model(input_size=(256, 256, 3)): 48 | inputs = Input(input_size) 49 | x = Conv2D(16, kernel_size=(3, 3), strides=(2, 2), padding='same', use_bias=True)(inputs) 50 | 51 | # block 1 ~ 9 (1, 128, 128, 16) 52 | x = conv_blocks(x, 16, activation='prelu', num_iterations=8) 53 | x = conv_blocks_with_pooling(x, 32, activation='prelu', padding=True, pad_value=16) 54 | 55 | # block 10 ~ 18 (1, 64, 64, 32) 56 | x = conv_blocks(x, 32, activation='prelu', num_iterations=8) 57 | x = conv_blocks_with_pooling(x, 64, activation='prelu', padding=True, pad_value=32) 58 | 59 | # block 19 ~ 27 (1, 32, 32, 64) 60 | x = conv_blocks(x, 64, activation='prelu', num_iterations=8) 61 | x = PReLU(alpha_initializer='uniform', shared_axes=[1, 2])(x) 62 | shortcut = x 63 | shortcut = MaxPooling2D(pool_size=(2, 2), strides=(2, 2), padding='valid')(shortcut) 64 | paddings = tf.constant([[0,0],[0,0],[0,0],[0,192]]) 65 | shortcut = Lambda(lambda x: tf.pad(x, paddings, "CONSTANT"))(shortcut) 66 | x = Conv2D(128, kernel_size=(2, 2), strides=(2, 2), padding='valid', use_bias=True)(x) 67 | x = PReLU(alpha_initializer='uniform', shared_axes=[1, 2])(x) 68 | x = DepthwiseConv2D(kernel_size=(3, 3), strides=(1, 1), padding='same', use_bias=True)(x) 69 | x = Conv2D(256, kernel_size=(1, 1), strides=(1, 1), padding='valid', use_bias=True)(x) 70 | x = Add()([x, shortcut]) 71 | 72 | # block 28 ~ 36 (1, 16, 16, 256) 73 | x = conv_blocks(x, 256, activation='prelu', num_iterations=8) 74 | x = conv_blocks_with_pooling(x, 256, activation='prelu', padding=False) 75 | 76 | # block 37 ~ 45 (1, 8, 8, 256) 77 | x = conv_blocks(x, 256, activation='prelu', num_iterations=8) 78 | x = conv_blocks_with_pooling(x, 256, activation='prelu', padding=False) 79 | 80 | # block 46 ~ 54 (1, 4, 4, 256) 81 | x = conv_blocks(x, 256, activation='prelu', num_iterations=8) 82 | x = conv_blocks_with_pooling(x, 256, activation='prelu', padding=False) 83 | 84 | # block 55 ~ 63 (1, 2, 2, 256) 85 | x = conv_blocks(x, 256, activation='prelu', num_iterations=8) 86 | 87 | # Last layer (1, 2, 2, 256) 88 | x = PReLU(alpha_initializer='uniform', shared_axes=[1, 2])(x) 89 | ## Hand_flag value (1, 1) 90 | hand_flag = Conv2D(1, kernel_size=(2, 2), strides=(1, 1), padding='valid', use_bias=True, name='conv_handflag')(x) 91 | hand_flag = Activation('sigmoid', name='activation_handflag')(hand_flag) 92 | hand_flag = Reshape(target_shape=(1,), name='output_handflag')(hand_flag) 93 | ## Hand_Landmark_2d (1, 42) 94 | landmarks = Conv2D(42, kernel_size=(2, 2), strides=(1, 1), padding='valid', use_bias=True, name='convld_21_2d')(x) 95 | landmarks = Reshape(target_shape=(42,), name='ld_21_2d')(landmarks) 96 | 97 | model = Model(inputs, [landmarks, hand_flag]) 98 | return model 99 | -------------------------------------------------------------------------------- /hand_landmark/model_3d.py: -------------------------------------------------------------------------------- 1 | from tensorflow.keras.layers import Input, Conv2D, PReLU, DepthwiseConv2D, Add, ReLU, MaxPooling2D, Reshape, Lambda, Activation 2 | from tensorflow.keras.activations import sigmoid 3 | from tensorflow.keras.models import Model 4 | import tensorflow as tf 5 | 6 | # Mediapipe: hand_landmark_3d.tflite 7 | def conv_blocks(x, num_filter, activation='relu', num_iterations=1): 8 | for num_iter in range(0, num_iterations): 9 | if activation == 'relu': 10 | x = ReLU()(x) 11 | elif activation == 'prelu': 12 | x = PReLU(alpha_initializer='uniform', shared_axes=[1, 2])(x) 13 | shortcut = x 14 | x = DepthwiseConv2D(kernel_size=(5, 5), strides=(1, 1), padding='same', use_bias=True)(x) 15 | x = Conv2D(num_filter, kernel_size=(1, 1), strides=(1, 1), padding='valid', use_bias=True)(x) 16 | x = Add()([shortcut, x]) 17 | return x 18 | 19 | def conv_blocks_with_pooling(x, num_filter, activation='prelu', padding=False, pad_value=None): 20 | if activation == 'relu': 21 | x = ReLU()(x) 22 | elif activation == 'prelu': 23 | x = PReLU(alpha_initializer='uniform', shared_axes=[1, 2])(x) 24 | 25 | shortcut = x 26 | shortcut = MaxPooling2D(pool_size=(2, 2), strides=(2, 2), padding='valid')(shortcut) 27 | 28 | if padding: 29 | paddings = tf.constant([[0,0],[0,0],[0,0],[0,pad_value]]) 30 | shortcut = Lambda(lambda x: tf.pad(x, paddings, "CONSTANT"))(shortcut) # CONSTANT / REFLECT / SYMMETRIC 31 | x = DepthwiseConv2D(kernel_size=(5, 5), strides=(2, 2), padding='same', use_bias=True)(x) 32 | x = Conv2D(num_filter, kernel_size=(1, 1), strides=(1, 1), padding='valid', use_bias=True)(x) 33 | x = Add()([shortcut, x]) 34 | else: 35 | x = DepthwiseConv2D(kernel_size=(5, 5), strides=(2, 2), padding='same', use_bias=True)(x) 36 | x = Conv2D(num_filter, kernel_size=(1, 1), strides=(1, 1), padding='valid', use_bias=True)(x) 37 | x = Add()([shortcut, x]) 38 | return x 39 | 40 | def hand_landmark_3d_model(input_size=(256, 256, 3)): 41 | inputs = Input(input_size) 42 | x = Conv2D(32, kernel_size=(3, 3), strides=(2, 2), padding='same', use_bias=True)(inputs) 43 | 44 | # block 1 ~ 5 (1, 128, 128, 32) 45 | x = conv_blocks(x, 32, activation='prelu', num_iterations=1) 46 | x = conv_blocks(x, 32, activation='relu', num_iterations=3) 47 | x = conv_blocks_with_pooling(x, 64, activation='relu', padding=True, pad_value=32) 48 | 49 | # block 6 ~ 11 (1, 64, 64, 64) 50 | x = conv_blocks(x, 64, activation='relu', num_iterations=5) 51 | x = conv_blocks_with_pooling(x, 128, activation='relu', padding=True, pad_value=64) 52 | 53 | # block 12 ~ 18 (1, 32, 32, 128) 54 | x = conv_blocks(x, 128, activation='relu', num_iterations=6) 55 | x = conv_blocks_with_pooling(x, 192, activation='relu', padding=True, pad_value=64) 56 | 57 | # block 19 ~ 25 (1, 16, 16, 192) 58 | x = conv_blocks(x, 192, activation='relu', num_iterations=6) 59 | x = conv_blocks_with_pooling(x, 192, activation='relu', padding=False) 60 | 61 | # block 26 ~ 32 (1, 8, 8, 192) 62 | x = conv_blocks(x, 192, activation='relu', num_iterations=6) 63 | x = conv_blocks_with_pooling(x, 192, activation='relu', padding=False) 64 | 65 | # block 33 ~ 39 (1, 4, 4, 192) 66 | x = conv_blocks(x, 192, activation='relu', num_iterations=6) 67 | x = conv_blocks_with_pooling(x, 192, activation='relu', padding=False) 68 | 69 | # block 40 ~ 45 (1, 2, 2, 192) 70 | x = conv_blocks(x, 192, activation='relu', num_iterations=6) 71 | 72 | # Last layer 73 | x = ReLU()(x) 74 | ## output_handflag 75 | hand_flag = Conv2D(1, kernel_size=(2, 2), strides=(1, 1), padding='valid', use_bias=True, name='conv_handflag')(x) 76 | hand_flag = Activation('sigmoid', name='activation_handflag')(hand_flag) 77 | hand_flag = Reshape(target_shape=(1,), name='output_handflag')(hand_flag) 78 | 79 | ## Hand_Landmark_3d (1, 63) 80 | landmarks = Conv2D(63, kernel_size=(2, 2), strides=(1, 1), padding='valid', use_bias=True, name='convld_21_3d')(x) 81 | landmarks = Reshape(target_shape=(63,), name='ld_21_3d')(landmarks) 82 | 83 | model = Model(inputs, [hand_flag, landmarks]) 84 | return model 85 | -------------------------------------------------------------------------------- /hand_landmark/utils.py: -------------------------------------------------------------------------------- 1 | import tensorflow as tf 2 | import numpy as np 3 | import tensorflow.keras.backend as K 4 | 5 | def get_pretrained_tflite_weights(model_path): 6 | model = tf.lite.Interpreter(model_path) 7 | tensor_details = model.get_tensor_details() 8 | weights_dict = {} 9 | layer_names = [] 10 | 11 | for idx in range(0, len(tensor_details)): 12 | try: 13 | name = tensor_details[idx]['name'] 14 | weights = model.get_tensor(idx) 15 | weights_dict[name] = weights 16 | except: 17 | name = tensor_details[idx]['name'] 18 | layer_names.append(name) 19 | return weights_dict, layer_names 20 | 21 | def set_pretrained_weights(model, weights_dict, layer_names): 22 | for name in layer_names: 23 | if name.find('conv') != -1: 24 | pretrained_weights = [] 25 | kernel_weight = weights_dict.get(name+"/Kernel") 26 | bias_weight = weights_dict.get(name+'/Bias') 27 | kernel_weight = kernel_weight.transpose(1, 2, 3, 0) 28 | pretrained_weights.append(kernel_weight) 29 | pretrained_weights.append(bias_weight) 30 | layer = model.get_layer(name) 31 | layer.set_weights(pretrained_weights) 32 | 33 | elif name.find('p_re_lu') != -1: 34 | pretrained_weights = [] 35 | alpha_weight = weights_dict.get(name+"/Alpha") 36 | pretrained_weights.append(alpha_weight) 37 | 38 | layer = model.get_layer(name) 39 | layer.set_weights(pretrained_weights) 40 | 41 | def define_fake_4_channels_graph(input_size=(256, 256, 4)): 42 | inputs = tf.keras.layers.Input(input_size) 43 | x = tf.keras.layers.Lambda(lambda x: x[:,:,:,:3], name='slicing_inputs')(inputs) # slicing into 3-channels 44 | model = tf.keras.models.Model(inputs, x) 45 | sess = K.get_session() 46 | fake_graph = sess.graph_def 47 | K.clear_session() 48 | return fake_graph 49 | 50 | def display_nodes(nodes): 51 | for i, node in enumerate(nodes): 52 | print("%d %s %s" % (i, node.name, node.op)) 53 | for idx, n in enumerate(node.input): 54 | print(u"└─── %d ─ %s" % (idx, n)) 55 | -------------------------------------------------------------------------------- /palm_detection/assets/palm_7_landmark_index.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/assets/palm_7_landmark_index.png -------------------------------------------------------------------------------- /palm_detection/mediapipe_models/palm_detection.tflite: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/mediapipe_models/palm_detection.tflite -------------------------------------------------------------------------------- /palm_detection/mediapipe_models/palm_detection_4channels.pb: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/mediapipe_models/palm_detection_4channels.pb -------------------------------------------------------------------------------- /palm_detection/mediapipe_models/palm_detection_4channels.tflite: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/mediapipe_models/palm_detection_4channels.tflite -------------------------------------------------------------------------------- /palm_detection/mediapipe_models/palm_detection_without_custom_layer.tflite: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/mediapipe_models/palm_detection_without_custom_layer.tflite -------------------------------------------------------------------------------- /palm_detection/model.py: -------------------------------------------------------------------------------- 1 | from tensorflow.keras.layers import Input, Conv2D, Conv2DTranspose, Add, ReLU, MaxPooling2D, Reshape, Lambda, Activation, DepthwiseConv2D, Conv2D, Concatenate 2 | from tensorflow.keras.models import Model 3 | import tensorflow as tf 4 | import numpy as np 5 | 6 | def conv_blocks(x, num_filter, num_iterations=1): 7 | for num_iter in range(0, num_iterations): 8 | x = ReLU()(x) 9 | shortcut = x 10 | x = DepthwiseConv2D(kernel_size=(3, 3), strides=(1, 1), padding='same', use_bias=True)(x) 11 | x = Conv2D(num_filter, kernel_size=(1, 1), strides=(1, 1), padding='valid', use_bias=True)(x) 12 | x = Add()([shortcut, x]) 13 | return x 14 | 15 | def conv_blocks_with_pooling(x, num_filter, padding=False, pad_value=None): 16 | x = ReLU()(x) 17 | shortcut = x 18 | shortcut = MaxPooling2D(pool_size=(2, 2), strides=(2, 2), padding='same')(shortcut) 19 | 20 | if padding: 21 | paddings = tf.constant([[0,0],[0,0],[0,0],[0,pad_value]]) 22 | shortcut = Lambda(lambda x: tf.pad(x, paddings, "CONSTANT"))(shortcut) # CONSTANT / REFLECT / SYMMETRIC 23 | x = DepthwiseConv2D(kernel_size=(3, 3), strides=(2, 2), padding='same', use_bias=True)(x) 24 | x = Conv2D(num_filter, kernel_size=(1, 1), strides=(1, 1), padding='valid', use_bias=True)(x) 25 | x = Add()([x, shortcut]) 26 | else: 27 | x = DepthwiseConv2D(kernel_size=(3, 3), strides=(2, 2), padding='same', use_bias=True)(x) 28 | x = Conv2D(num_filter, kernel_size=(1, 1), strides=(1, 1), padding='valid', use_bias=True)(x) 29 | x = Add()([x, shortcut]) 30 | return x 31 | 32 | def palm_detection_model(input_size=(256, 256, 3)): 33 | inputs = Input(input_size) 34 | x = Conv2D(32, kernel_size=(3, 3), strides=(2, 2), padding='same', use_bias=True)(inputs) 35 | 36 | # block 1 ~ 8 (1, 128, 128, 32) 37 | x = conv_blocks(x, 32, num_iterations=7) 38 | x = conv_blocks_with_pooling(x, 64, padding=True, pad_value=32) 39 | 40 | # block 9 ~ 16 (1, 64, 64, 64) 41 | x = conv_blocks(x, 64, num_iterations=7) 42 | x = conv_blocks_with_pooling(x, 128, padding=True, pad_value=64) 43 | 44 | # block 17 ~ 24 (1, 32, 32, 128) 45 | x = conv_blocks(x, 128, num_iterations=7) 46 | x = ReLU()(x) 47 | shortcut_1 = x #(1, 32, 32, 128) 48 | 49 | shortcut = x 50 | shortcut = MaxPooling2D(pool_size=(2, 2), strides=(2, 2), padding='same')(shortcut) 51 | paddings = tf.constant([[0,0],[0,0],[0,0],[0,128]]) 52 | shortcut = Lambda(lambda x: tf.pad(x, paddings, "CONSTANT"))(shortcut) # CONSTANT / REFLECT / SYMMETRIC 53 | x = DepthwiseConv2D(kernel_size=(3, 3), strides=(2, 2), padding='same', use_bias=True)(x) 54 | x = Conv2D(256, kernel_size=(1, 1), strides=(1, 1), padding='valid', use_bias=True)(x) 55 | x = Add()([x, shortcut]) 56 | 57 | # block 25 ~ 32 (1, 16, 16, 256) 58 | x = conv_blocks(x, 256, num_iterations=7) 59 | x = ReLU()(x) 60 | shortcut_2 = x #(1, 16, 16, 256) 61 | 62 | shortcut = x 63 | shortcut = MaxPooling2D(pool_size=(2, 2), strides=(2, 2), padding='same')(shortcut) 64 | x = DepthwiseConv2D(kernel_size=(3, 3), strides=(2, 2), padding='same', use_bias=True)(x) 65 | x = Conv2D(256, kernel_size=(1, 1), strides=(1, 1), padding='valid', use_bias=True)(x) 66 | x = Add()([x, shortcut]) 67 | 68 | # block 33 ~ 39 (1, 8, 8, 256) 69 | x = conv_blocks(x, 256, num_iterations=7) 70 | 71 | # Last layers 72 | ## (1, 8, 8, 256) 73 | x = ReLU()(x) 74 | shortcut_3 = x # (1, 8, 8, 256) 75 | x = Conv2DTranspose(256, kernel_size=(2, 2), strides=(2, 2), padding='same', use_bias=True)(x) 76 | x = ReLU()(x) 77 | x = Add()([x, shortcut_2]) 78 | 79 | ## (1, 16, 16, 256) 80 | shortcut = x 81 | x = DepthwiseConv2D(kernel_size=(3, 3), strides=(1, 1), padding='same', use_bias=True)(x) 82 | x = Conv2D(256, kernel_size=(1, 1), strides=(1, 1), padding='valid', use_bias=True)(x) 83 | x = Add()([x, shortcut]) 84 | 85 | x = ReLU()(x) 86 | shortcut_4 = x # (1, 16, 16, 256) 87 | x = Conv2DTranspose(128, kernel_size=(2, 2), strides=(2, 2), padding='same', use_bias=True)(x) 88 | x = ReLU()(x) 89 | x = Add()([x, shortcut_1]) 90 | 91 | ## (1, 32, 32, 128) 92 | shortcut = x 93 | x = DepthwiseConv2D(kernel_size=(3, 3), strides=(1, 1), padding='same', use_bias=True)(x) 94 | x = Conv2D(128, kernel_size=(1, 1), strides=(1, 1), padding='valid', use_bias=True)(x) 95 | x = Add()([x, shortcut]) 96 | x = ReLU()(x) 97 | 98 | ### Last block - Classificators 99 | classificator_8 = Conv2D(2, kernel_size=(1, 1), strides=(1, 1), padding='same', use_bias=True, name='classificator_8')(x) 100 | classificator_16 = Conv2D(2, kernel_size=(1, 1), strides=(1, 1), padding='same', use_bias=True, name='classificator_16')(shortcut_4) 101 | classificator_32 = Conv2D(6, kernel_size=(1, 1), strides=(1, 1), padding='same', use_bias=True, name='classificator_32')(shortcut_3) 102 | #### Reshape and Concatenation 103 | classificator_8 = Reshape(target_shape=(-1, 1))(classificator_8) # (1, 2048, 1) 104 | classificator_16 = Reshape(target_shape=(-1, 1))(classificator_16) # (1, 512, 1) 105 | classificator_32 = Reshape(target_shape=(-1, 1))(classificator_32) # (1, 384, 1) 106 | classificator_concat = Concatenate(axis=1, name='classificators')([classificator_8, classificator_16, classificator_32]) # (1, 2944, 1) 107 | 108 | ### Last block - Regressors 109 | regressor_8 = Conv2D(36, kernel_size=(1, 1), strides=(1, 1), padding='same', use_bias=True, name='regressor_8')(x) 110 | regressor_16 = Conv2D(36, kernel_size=(1, 1), strides=(1, 1), padding='same', use_bias=True, name='regressor_16')(shortcut_4) 111 | regressor_32 = Conv2D(108, kernel_size=(1, 1), strides=(1, 1), padding='same', use_bias=True, name='regressor_32')(shortcut_3) 112 | #### Reshape and Concatenation 113 | regressor_8 = Reshape(target_shape=(-1, 18))(regressor_8) # (1, 2048, 18) 114 | regressor_16 = Reshape(target_shape=(-1, 18))(regressor_16) # (1, 512, 18) 115 | regressor_32 = Reshape(target_shape=(-1, 18))(regressor_32) # (1, 384, 18) 116 | regressor_concat = Concatenate(axis=1, name='regressors')([regressor_8, regressor_16, regressor_32]) 117 | 118 | model = Model(inputs, [regressor_concat, classificator_concat]) 119 | return model 120 | -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/classificator_16_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/classificator_16_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/classificator_16_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/classificator_16_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/classificator_32_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/classificator_32_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/classificator_32_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/classificator_32_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/classificator_8_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/classificator_8_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/classificator_8_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/classificator_8_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_10_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_10_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_10_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_10_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_11_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_11_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_11_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_11_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_12_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_12_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_12_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_12_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_13_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_13_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_13_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_13_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_14_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_14_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_14_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_14_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_15_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_15_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_15_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_15_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_16_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_16_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_16_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_16_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_17_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_17_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_17_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_17_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_18_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_18_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_18_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_18_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_19_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_19_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_19_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_19_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_1_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_1_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_1_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_1_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_20_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_20_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_20_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_20_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_21_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_21_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_21_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_21_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_22_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_22_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_22_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_22_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_23_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_23_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_23_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_23_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_24_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_24_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_24_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_24_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_25_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_25_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_25_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_25_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_26_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_26_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_26_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_26_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_27_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_27_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_27_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_27_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_28_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_28_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_28_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_28_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_29_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_29_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_29_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_29_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_2_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_2_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_2_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_2_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_30_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_30_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_30_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_30_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_31_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_31_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_31_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_31_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_32_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_32_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_32_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_32_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_33_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_33_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_33_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_33_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_34_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_34_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_34_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_34_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_35_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_35_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_35_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_35_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_36_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_36_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_36_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_36_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_37_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_37_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_37_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_37_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_38_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_38_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_38_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_38_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_39_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_39_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_39_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_39_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_3_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_3_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_3_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_3_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_40_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_40_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_40_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_40_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_41_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_41_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_41_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_41_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_4_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_4_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_4_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_4_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_5_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_5_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_5_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_5_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_6_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_6_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_6_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_6_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_7_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_7_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_7_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_7_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_8_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_8_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_8_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_8_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_9_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_9_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_9_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_9_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_transpose_1_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_transpose_1_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_transpose_1_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_transpose_1_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_transpose_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_transpose_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/conv2d_transpose_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/conv2d_transpose_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_10_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_10_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_10_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_10_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_11_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_11_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_11_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_11_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_12_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_12_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_12_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_12_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_13_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_13_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_13_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_13_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_14_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_14_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_14_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_14_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_15_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_15_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_15_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_15_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_16_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_16_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_16_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_16_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_17_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_17_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_17_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_17_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_18_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_18_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_18_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_18_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_19_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_19_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_19_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_19_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_1_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_1_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_1_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_1_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_20_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_20_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_20_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_20_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_21_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_21_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_21_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_21_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_22_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_22_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_22_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_22_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_23_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_23_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_23_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_23_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_24_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_24_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_24_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_24_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_25_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_25_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_25_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_25_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_26_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_26_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_26_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_26_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_27_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_27_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_27_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_27_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_28_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_28_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_28_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_28_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_29_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_29_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_29_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_29_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_2_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_2_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_2_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_2_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_30_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_30_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_30_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_30_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_31_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_31_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_31_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_31_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_32_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_32_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_32_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_32_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_33_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_33_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_33_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_33_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_34_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_34_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_34_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_34_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_35_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_35_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_35_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_35_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_36_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_36_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_36_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_36_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_37_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_37_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_37_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_37_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_38_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_38_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_38_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_38_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_39_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_39_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_39_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_39_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_3_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_3_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_3_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_3_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_40_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_40_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_40_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_40_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_4_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_4_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_4_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_4_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_5_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_5_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_5_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_5_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_6_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_6_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_6_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_6_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_7_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_7_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_7_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_7_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_8_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_8_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_8_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_8_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_9_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_9_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_9_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_9_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/depthwise_conv2d_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/depthwise_conv2d_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/regressor_16_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/regressor_16_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/regressor_16_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/regressor_16_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/regressor_32_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/regressor_32_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/regressor_32_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/regressor_32_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/regressor_8_Bias.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/regressor_8_Bias.npy -------------------------------------------------------------------------------- /palm_detection/pretrained_weights/regressor_8_Kernel.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/junhwanjang/mediapipe-models/3c979eb59389dab4e0d1e077a0f5cf2d08bb3582/palm_detection/pretrained_weights/regressor_8_Kernel.npy -------------------------------------------------------------------------------- /palm_detection/utils.py: -------------------------------------------------------------------------------- 1 | import tensorflow as tf 2 | import numpy as np 3 | import tensorflow.keras.backend as K 4 | 5 | def define_fake_4_channels_graph(input_size=(256, 256, 4)): 6 | inputs = tf.keras.layers.Input(input_size) 7 | x = tf.keras.layers.Lambda(lambda x: x[:,:,:,:3], name='slicing_inputs')(inputs) # slicing into 3-channels 8 | model = tf.keras.models.Model(inputs, x) 9 | sess = K.get_session() 10 | fake_graph = sess.graph_def 11 | K.clear_session() 12 | return fake_graph 13 | 14 | def set_pretrained_weights(model, weights_dir='pretrained_weights/'): 15 | # Get layer names 16 | layer_names = [ 17 | 'classificator_8', 'classificator_16', 'classificator_32', 18 | 'regressor_8', 'regressor_16', 'regressor_32', 19 | 'conv2d', 'depthwise_conv2d', 'conv2d_transpose', 'conv2d_transpose_1' 20 | ] 21 | num_conv = 41 22 | num_depth_conv = 40 23 | # Append conv2d layer names 24 | for i in range(1, num_conv+1): 25 | name = 'conv2d_' + str(i) 26 | layer_names.append(name) 27 | for i in range(1, num_depth_conv+1): 28 | name = 'depthwise_conv2d_' + str(i) 29 | layer_names.append(name) 30 | layer_names.sort() 31 | 32 | # Set pretrained weights from npy file 33 | for name in layer_names: 34 | pretrained_weights = [] 35 | kernel_weight_path = weights_dir + name + "_Kernel.npy" 36 | bias_weight_path = weights_dir + name + "_Bias.npy" 37 | kernel_weight = np.load(kernel_weight_path) 38 | bias_weight = np.load(bias_weight_path) 39 | if name.find("conv2d_transpose") == -1: 40 | kernel_weight = kernel_weight.transpose(1, 2, 3, 0) 41 | else: 42 | kernel_weight = kernel_weight.transpose(1, 2, 0, 3) 43 | 44 | pretrained_weights.append(kernel_weight) 45 | pretrained_weights.append(bias_weight) 46 | layer = model.get_layer(name) 47 | layer.set_weights(pretrained_weights) 48 | 49 | print("[INFO] Set all pretrained weights") 50 | 51 | def display_nodes(nodes): 52 | for i, node in enumerate(nodes): 53 | print('%d %s %s' % (i, node.name, node.op)) 54 | for idx, n in enumerate(node.input): 55 | print(u'└─── %d ─ %s' % (idx, n)) 56 | --------------------------------------------------------------------------------