{"id":13563705,"url":"https://github.com/hongfz16/AvatarCLIP","last_synced_at":"2025-04-03T20:31:50.234Z","repository":{"id":37683484,"uuid":"488129853","full_name":"hongfz16/AvatarCLIP","owner":"hongfz16","description":"[SIGGRAPH 2022 Journal Track] AvatarCLIP: Zero-Shot Text-Driven Generation and Animation of 3D Avatars","archived":false,"fork":false,"pushed_at":"2023-02-15T12:46:00.000Z","size":22462,"stargazers_count":1081,"open_issues_count":19,"forks_count":98,"subscribers_count":21,"default_branch":"main","last_synced_at":"2025-04-03T14:11:03.597Z","etag":null,"topics":["3d-generation","motion-generation","nerf"],"latest_commit_sha":null,"homepage":"https://hongfz16.github.io/projects/AvatarCLIP.html","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"other","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/hongfz16.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null}},"created_at":"2022-05-03T08:23:30.000Z","updated_at":"2025-03-29T20:07:41.000Z","dependencies_parsed_at":"2024-01-20T19:07:14.111Z","dependency_job_id":"5c461843-97fa-445d-9d6e-2884f6145cb1","html_url":"https://github.com/hongfz16/AvatarCLIP","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/hongfz16%2FAvatarCLIP","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/hongfz16%2FAvatarCLIP/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/hongfz16%2FAvatarCLIP/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/hongfz16%2FAvatarCLIP/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/hongfz16","download_url":"https://codeload.github.com/hongfz16/AvatarCLIP/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":247074885,"owners_count":20879342,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["3d-generation","motion-generation","nerf"],"created_at":"2024-08-01T13:01:22.431Z","updated_at":"2025-04-03T20:31:50.225Z","avatar_url":"https://github.com/hongfz16.png","language":"Python","funding_links":[],"categories":["Python","Papers"],"sub_categories":["Text-Driven motion generation"],"readme":"\u003cdiv align=\"center\"\u003e\n\n\u003ch1\u003eAvatarCLIP: Zero-Shot Text-Driven Generation and Animation of 3D Avatars\u003c/h1\u003e\n\n\u003cdiv\u003e\n    \u003ca href='https://hongfz16.github.io/' target='_blank'\u003eFangzhou Hong\u003c/a\u003e\u003csup\u003e1\u003c/sup\u003e*\u0026emsp;\n    \u003ca href='https://scholar.google.com/citations?user=2QLD4fAAAAAJ\u0026hl=en' target='_blank'\u003eMingyuan Zhang\u003c/a\u003e\u003csup\u003e1\u003c/sup\u003e*\u0026emsp;\n    \u003ca href='https://scholar.google.com/citations?user=lSDISOcAAAAJ\u0026hl=zh-CN' target='_blank'\u003eLiang Pan\u003c/a\u003e\u003csup\u003e1\u003c/sup\u003e\u0026emsp;\n    \u003ca href='https://caizhongang.github.io/' target='_blank'\u003eZhongang Cai\u003c/a\u003e\u003csup\u003e1,2,3\u003c/sup\u003e\u0026emsp;\n    \u003ca href='https://yanglei.me/' target='_blank'\u003eLei Yang\u003c/a\u003e\u003csup\u003e2\u003c/sup\u003e\u0026emsp;\n    \u003ca href='https://liuziwei7.github.io/' target='_blank'\u003eZiwei Liu\u003c/a\u003e\u003csup\u003e1+\u003c/sup\u003e\n\u003c/div\u003e\n\u003cdiv\u003e\n    \u003csup\u003e1\u003c/sup\u003eS-Lab, Nanyang Technological University\u0026emsp;\n    \u003csup\u003e2\u003c/sup\u003eSenseTime Research\u0026emsp;\n    \u003csup\u003e3\u003c/sup\u003eShanghai AI Laboratory\n\u003c/div\u003e\n\u003cdiv\u003e\n    *equal contribution\u0026emsp;\n    \u003csup\u003e+\u003c/sup\u003ecorresponding author\n\u003c/div\u003e\n\n\u003cstrong\u003eAccepted to \u003ca href='https://s2022.siggraph.org/' target='_blank'\u003eSIGGRAPH 2022\u003c/a\u003e (Journal Track)\u003c/strong\u003e\n\n\u003ch3\u003eTL;DR\u003c/h3\u003e\n\u003ch4\u003eAvatarCLIP generate and animate avatars given descriptions of \u003cspan style=\"color:#0a939d\"\u003ebody shapes\u003c/span\u003e, \u003cspan style=\"color:#EE9B00\"\u003eappearances\u003c/span\u003e and \u003cspan style=\"color:#AE2011\"\u003emotions\u003c/span\u003e.\u003c/h4\u003e\n\n\u003ctable\u003e\n\u003ctr\u003e\n    \u003ctd\u003e\u003cimg src=\"assets/tallandskinny_femalesoldier_arguing.gif\" width=\"100%\"/\u003e\u003c/td\u003e\n    \u003ctd\u003e\u003cimg src=\"assets/skinny_ninja_raisingbotharms.gif\" width=\"100%\"/\u003e\u003c/td\u003e\n    \u003ctd\u003e\u003cimg src=\"assets/overweight_sumowrestler_sitting.gif\" width=\"100%\"/\u003e\u003c/td\u003e\n    \u003ctd\u003e\u003cimg src=\"assets/tallandfat_ironman_running.gif\" width=\"100%\"/\u003e\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n    \u003ctd align='center' width='24%'\u003eA \u003cspan style=\"color:#0a939d\"\u003etall and skinny\u003c/span\u003e \u003cspan style=\"color:#EE9B00\"\u003efemale soldier\u003c/span\u003e that is \u003cspan style=\"color:#AE2011\"\u003earguing\u003c/span\u003e.\u003c/td\u003e\n    \u003ctd align='center' width='24%'\u003eA \u003cspan style=\"color:#0a939d\"\u003eskinny\u003c/span\u003e \u003cspan style=\"color:#EE9B00\"\u003eninja\u003c/span\u003e that is \u003cspan style=\"color:#AE2011\"\u003eraising both arms\u003c/span\u003e.\u003c/td\u003e\n    \u003ctd align='center' width='24%'\u003eAn \u003cspan style=\"color:#0a939d\"\u003eoverweight\u003c/span\u003e \u003cspan style=\"color:#EE9B00\"\u003esumo wrestler\u003c/span\u003e that is \u003cspan style=\"color:#AE2011\"\u003esitting\u003c/span\u003e.\u003c/td\u003e\n    \u003ctd align='center' width='24%'\u003eA \u003cspan style=\"color:#0a939d\"\u003etall and fat\u003c/span\u003e \u003cspan style=\"color:#EE9B00\"\u003eIron Man\u003c/span\u003e that is \u003cspan style=\"color:#AE2011\"\u003erunning\u003c/span\u003e.\u003c/td\u003e\n\u003ctr\u003e\n\u003c/table\u003e\n\nThis repository contains the official implementation of _AvatarCLIP: Zero-Shot Text-Driven Generation and Animation of 3D Avatars_.\n\n---\n\n\u003ch4 align=\"center\"\u003e\n  \u003ca href=\"https://hongfz16.github.io/projects/AvatarCLIP.html\" target='_blank'\u003e[Project Page]\u003c/a\u003e •\n  \u003ca href=\"https://arxiv.org/abs/2205.08535\" target='_blank'\u003e[arXiv]\u003c/a\u003e •\n  \u003ca href=\"https://drive.google.com/file/d/1_-5JIWyRCF7osAVWQ-z01nme4NxBTtrC/view?usp=sharing\" target='_blank'\u003e[High-Res PDF (166M)]\u003c/a\u003e •\n  \u003ca href=\"https://youtu.be/-l2ZMeoASGY\" target='_blank'\u003e[Supplementary Video]\u003c/a\u003e •\n  \u003ca href=\"https://colab.research.google.com/drive/1dfaecX7xF3nP6fyXc8XBljV5QY1lc1TR?usp=sharing\" target='_blank'\u003e[Colab Demo]\u003c/a\u003e\n\u003c/h4\u003e\n\n\u003c/div\u003e\n\n## Updates\n[09/2022] :fire::fire::fire:**If you are looking for a higher-quality 3D human generation method, go checkout our new work [EVA3D](https://hongfz16.github.io/projects/EVA3D.html)!**:fire::fire::fire:\n\n[09/2022] :fire::fire::fire:**If you are looking for a higher-quality text2motion method, go checkout our new work [MotionDiffuse](https://mingyuan-zhang.github.io/projects/MotionDiffuse.html)!**:fire::fire::fire:\n\n[07/2022] Code release for motion generation part!\n\n[05/2022] [Paper](https://arxiv.org/abs/2205.08535) uploaded to arXiv. [![arXiv](https://img.shields.io/badge/arXiv-2205.08535-b31b1b.svg)](https://arxiv.org/abs/2205.08535)\n\n[05/2022] Add a [Colab Demo](https://colab.research.google.com/drive/1dfaecX7xF3nP6fyXc8XBljV5QY1lc1TR?usp=sharing) for avatar generation! [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/1dfaecX7xF3nP6fyXc8XBljV5QY1lc1TR?usp=sharing)\n\n[05/2022] Support converting the generated avatar to the **animatable FBX format**! Go checkout [how to use the FBX models](#use-generated-fbx-models). Or checkout the [instructions](./Avatar2FBX/README.md) for the conversion codes.\n\n[05/2022] Code release for avatar generation part!\n\n[04/2022] AvatarCLIP is accepted to SIGGRAPH 2022 (Journal Track):partying_face:!\n\n## Citation\nIf you find our work useful for your research, please consider citing the paper:\n```\n@article{hong2022avatarclip,\n    title={AvatarCLIP: Zero-Shot Text-Driven Generation and Animation of 3D Avatars},\n    author={Hong, Fangzhou and Zhang, Mingyuan and Pan, Liang and Cai, Zhongang and Yang, Lei and Liu, Ziwei},\n    journal={ACM Transactions on Graphics (TOG)},\n    volume={41},\n    number={4},\n    articleno={161},\n    pages={1--19},\n    year={2022},\n    publisher={ACM New York, NY, USA},\n    doi={10.1145/3528223.3530094},\n}\n```\n\n## Use Generated FBX Models\n\n### Download\n\nGo visit our [project page](https://hongfz16.github.io/projects/AvatarCLIP.html). Go to the section 'Avatar Gallery'. Pick a model you like. Click 'Load Model' below. Click 'Download FBX' link at the bottom of the pop-up viewer.\n\n\u003cimg src='./assets/download_fbx.jpg' width='60%'\u003e\n\n### Import to Your Favourite 3D Software (e.g. Blender, Unity3D)\n\nThe FBX models are already rigged. Use your motion library to animate it!\n\n\u003cimg src='./assets/blender_tpose.png' width='60%'\u003e\n\n\u003cimg src='./assets/blender_motion.gif' width='100%'\u003e\n\n### Upload to Mixamo\n\nTo make use of the rich motion library provided by [Mixamo](https://www.mixamo.com), you can also upload the FBX model to Mixamo. The rigging process is completely automatic!\n\n\u003cimg src='./assets/mixamo_motion.gif' width='100%'\u003e\n\n## Installation\n\nWe recommend using anaconda to manage the python environment. The setup commands below are provided for your reference.\n\n```bash\ngit clone https://github.com/hongfz16/AvatarCLIP.git\ncd AvatarCLIP\nconda create -n AvatarCLIP python=3.7\nconda activate AvatarCLIP\nconda install pytorch==1.7.0 torchvision==0.8.0 torchaudio==0.7.0 cudatoolkit=10.1 -c pytorch\npip install -r requirements.txt\n```\n\nOther than the above steps, you should also install [neural_renderer](https://github.com/daniilidis-group/neural_renderer) following its instructions. Before compiling neural_renderer (or after compiling should also be fine), remember to add the following three lines to `neural_renderer/perspective.py` after line 19.\n\n```python\nx[z\u003c=0] = 0\ny[z\u003c=0] = 0\nz[z\u003c=0] = 0\n```\n\nThis quick fix is for a rendering issue where objects behide the camera will also be rendered. Be careful when using this fixed version of neural_renderer on your other projects, because this fix will cause the rendering process not differentiable.\n\nTo support offscreen rendering for motion visualization, you should install osmesa library.\n\n```bash\nconda install -c menpo osmesa\n```\n\n## Data Preparation\n\n### Download SMPL Models\nRegister and download SMPL models [here](https://smpl.is.tue.mpg.de/). Put the downloaded models in the folder `smpl_models`. The folder structure should look like\n\n```\n./\n├── ...\n└── smpl_models/\n    ├── smpl/\n        ├── SMPL_FEMALE.pkl\n        ├── SMPL_MALE.pkl\n        └── SMPL_NEUTRAL.pkl\n```\n\n### Download Pretrained Models \u0026 Other Data\nThis download is only for coarse shape generation and motion generation. You can skip if you only want to use other parts. Download the pretrained weights and other required data [here](https://1drv.ms/u/s!AjLpFg-f48ljgZl9qpU7_6ZA9B7qwA?e=pPcHIG). Put them in the folder `AvatarGen` so that the folder structure should look like\n\n```\n./\n├── ...\n└── AvatarGen/\n    └── ShapeGen/\n        └── data/\n            ├── codebook.pth\n            ├── model_VAE_16.pth\n            ├── nongrey_male_0110.jpg\n            ├── smpl_uv.mtl\n            └── smpl_uv.obj\n```\n\n\nPretrained weights and human texture for motion generation can be downloaded [here](https://drive.google.com/drive/folders/1TSyeT8MwH5EVQRbNGRVkWsA4Y9Y6dRbk?usp=sharing). Note that the human texture we used to render poses is from [SURREAL dataset](https://www.di.ens.fr/willow/research/surreal/data/). Besides, you should download pretrained weights of [VPoser v2.0](https://smpl-x.is.tue.mpg.de/download.php). Put them in the folder `AvatarAnimate` so that the folder structure should look like\n\n```\n├── ...\n└── AvatarAnimate/\n    └── data/\n        ├── codebook.pth\n        ├── motion_vae.pth\n        ├── pose_realnvp.pth\n        ├── nongrey_male_0110.jpg\n        ├── smpl_uv.mtl\n        ├── smpl_uv.obj\n        └── vposer\n            ├── V02_05.log\n            ├── V02_05.yaml\n            └── snapshots\n                ├── V02_05_epoch=08_val_loss=0.03.ckpt\n                └── V02_05_epoch=13_val_loss=0.03.ckpt\n        \n```\n\n## Avatar Generation\n\n### Coarse Shape Generation\n\nFolder `AvatarGen/ShapeGen` contains codes for this part. Run the follow command to generate the coarse shape corresponding to the shape description 'a strong man'. We recommend to use the prompt augmentation 'a 3d rendering of xxx in unreal engine' for better results. The generated coarse body mesh will be stored under `AvatarGen/ShapeGen/output/coarse_shape`.\n\n```bash\npython main.py --target_txt 'a 3d rendering of a strong man in unreal engine'\n```\n\nThen we need to render the mesh for initialization of the implicit avatar representation. Use the following command for rendering.\n\n```bash\npython render.py --coarse_shape_obj output/coarse_shape/a_3d_rendering_of_a_strong_man_in_unreal_engine.obj --output_folder ${RENDER_FOLDER}\n```\n\n### Shape Sculpting and Texture Generation\n\nNote that all the codes are tested on NVIDIA V100 (32GB memory). Therefore, in order to run on GPUs with lower memory, please try to scale down the network or tune down `max_ray_num` in the config files. You can refer to `confs/examples_small/example.conf` or our [colab demo](https://colab.research.google.com/drive/1dfaecX7xF3nP6fyXc8XBljV5QY1lc1TR?usp=sharing) for a scale-down version of AvatarCLIP.\n\nFolder `AvatarGen/AppearanceGen` contains codes for this part. We provide data, pretrained model and scripts to perform shape sculpting and texture generation on a zero-beta body (mean shape defined by SMPL). We provide many example scripts under `AvatarGen/AppearanceGen/confs/examples`. For example, if we want to generate 'Abraham Lincoln', which is defined in the config file `confs/examples/abrahamlincoln.conf`, use the following command.\n\n```bash\npython main.py --mode train_clip --conf confs/examples/abrahamlincoln.conf\n```\n\nResults will be stored in `AvatarCLIP/AvatarGen/AppearanceGen/exp/smpl/examples/abrahamlincoln`.\n\nIf you wish to perform shape sculpting and texture generation on the previously generated coarse shape. We also provide example config files in `confs/base_models/astrongman.conf` `confs/astrongman/*.conf`. Two steps of optimization are required as follows.\n\n```bash\n# Initilization of the implicit avatar\npython main.py --mode train --conf confs/base_models/astrongman.conf\n# Shape sculpting and texture generation on the initialized implicit avatar\npython main.py --mode train_clip --conf confs/astrongman/hulk.conf\n```\n\n### Marching Cube\n\nTo extract meshes from the generated implicit avatar, one may use the following command.\n\n```bash\npython main.py --mode validate_mesh --conf confs/examples/abrahamlincoln.conf\n```\n\nThe final high resolution mesh will be stored as `AvatarCLIP/AvatarGen/AppearanceGen/exp/smpl/examples/abrahamlincoln/meshes/00030000.ply`\n\n## Convert Avatar to FBX Format\n\nFor the convenience of using the generated avatar with modern graphics pipeline, we also provide scripts to rig the avatar and convert to FBX format. See the instructions [here](./Avatar2FBX/README.md).\n\n\n## Motion Generation\n\n### Candidate Poses Generation\n\nHere we provide four different methods for pose generation.\n\n1. PoseOptimizer: directly optimize on SMPL theta\n\n2. VPoserOptimizer: optimize the latent space of VPoser\n\n3. VPoserRealNVP: get latent codes of VPoser from pretrained conditional RealNVP\n\n4. VPoserCodebook: select the most similar poses to the given text feature\n\n\nWe provide configurations to compare these methods. Here are some examples:\n\n```bash\n# Suppose your current location is `AvatarCLIP/AvatarAnimate`\n\n# Use PoseOptimizer method to generate poses for \"arguing\"\npython main.py --conf confs/pose_ablation/pose_optimizer/argue.conf\n# Results are stored in `AvatarCLIP/AvatarAnimate/exp/pose_ablation/pose_optimizer/argue` directory\n# candidate_0.jpg, candidate_1.jpg, ..., candidate_4.jpg are the top-5 poses\n# candidate_0.npy, candidate_1.npy, ..., candidate_4.npy are corresponding parameters\n\n# Use VPoserOptimizer method to generate poses for \"praying\"\npython main.py --conf confs/pose_ablation/vposer_optimizer/pray.conf\n# Results are stored in `AvatarCLIP/AvatarAnimate/exp/pose_ablation/vposer_optimizer/pray` directory\n\n# Use VPoserRealNVP method to generate poses for \"shooting a basketball\"\npython main.py --conf confs/pose_ablation/vposer_realnvp/shoot_basketball.conf\n# Results are stored in `AvatarCLIP/AvatarAnimate/exp/pose_ablation/vposer_realnvp/shoot_basketball` directory\n\n# Use VPoserCodebook method to generate poses for \"running\"\npython main.py --conf confs/pose_ablation/vposer_codebook/run.conf\n# Results are stored in `AvatarCLIP/AvatarAnimate/exp/pose_ablation/vposer_codebook/run` directory\n```\n\n### Motion Generation\n\nHere we provide three different methods for motion generation.\n\n1. MotionInterpolation: directly interpolate between given poses\n\n2. MotionOptimizer (baseline): optimize latent code of a pretrained VAE with a simple reconstruction loss\n\n3. MotionOptimizer (ours): optimize latent code of a pretrained VAE with weighted reconstruction loss, delta loss, and clip loss\n\n\n\nWe provide configurations to compare these methods. Here are some examples:\n\n```bash\n# Suppose your current location is `AvatarCLIP/AvatarAnimate`\n\n# Use MotionInterpolation method to generate motion for \"arguing\"\npython main.py --conf confs/motion_ablation/interpolation/argue.conf\n# Results are stored in `AvatarCLIP/AvatarAnimate/exp/motion_ablation/interpolation/argue` directory\n# candidate_0.jpg, candidate_1.jpg, ..., candidate_4.jpg are the top-5 poses\n# candidate_0.npy, candidate_1.npy, ..., candidate_4.npy are corresponding parameters\n# motion.mp4 is the generated motion\n# motion.npy is corresponding parameters\n\n# Use MotionOptimizer (baseline) method to generate motion for \"praying\"\npython main.py --conf confs/motion_ablation/baseline/pray.conf\n# Results are stored in `AvatarCLIP/AvatarAnimate/exp/motion_ablation/baseline/pray` directory\n\n# Use MotionOptimizer (ours) method to generate motion for \"shooting a basketball\"\npython main.py --conf confs/motion_ablation/motion_optimizer/shoot_basketball.conf\n# Results are stored in `AvatarCLIP/AvatarAnimate/exp/motion_ablation/motion_optimizer/shoot_basketball` directory\n```\n\n### Make your own configure\n\nEach configuration contains three independent parts: general setting, pose generator, and motion generator.\n\n```text\n# General Setting\ngeneral {\n    # describe the results path\n    base_exp_dir = ./exp/motion_ablation/motion_optimizer/raise_arms\n\n    # if you only want to generate poses, then you can set \"mode = pose\".\n    mode = motion\n\n    # define your prompt. We highly recommend using the format \"a rendered 3d man is xxx\"\n    text = a rendered 3d man is raising both arms\n}\n\n# Pose Generator\npose_generator {\n    type = VPoserCodebook\n    # you can change the number of candidate poses by setting \"topk = 10\"\n    # for PoseOptimizer and VPoserOptimizer, you can further define the number of iterations and the optimizer type\n}\n\n# Motion Generator\n# if \"mode = pose\", you can ignore this part\nmotion_generator {\n    type = MotionOptimizer\n    # you can further modify the coefficient of each loss. \n    # for example, if you find the generated motion is very intensive, you can reduce the coefficient of delta loss.\n}\n\n\n```\n\n\n\n## License\n\nDistributed under the S-Lab License. See `LICENSE` for more information.\n\n\n## Related Works\n\u003cp\u003eThere are lots of wonderful works that inspired our work or came around the same time as ours.\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://arxiv.org/abs/2112.01455\"\u003eDream Fields\u003c/a\u003e enables zero-shot text-driven general 3D object generation using CLIP and NeRF.\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://arxiv.org/abs/2112.03221\"\u003eText2Mesh\u003c/a\u003e proposes to edit a template mesh by predicting offsets and colors per vertex using CLIP and differentiable rendering.\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://arxiv.org/abs/2112.05139\"\u003eCLIP-NeRF\u003c/a\u003e can manipulate 3D objects represented by NeRF with natural languages or examplar images by leveraging CLIP.\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://arxiv.org/abs/2203.13333\"\u003eText to Mesh\u003c/a\u003e facilitates zero-shot text-driven general mesh generation by deforming from a sphere mesh guided by CLIP.\u003c/p\u003e\n\u003cp\u003e\u003ca href='https://github.com/GuyTevet/MotionCLIP'\u003eMotionCLIP\u003c/a\u003e establishes a projection from the CLIP text space to the motion space through supervised training, which leads to amazing text-driven motion generation results.\u003c/p\u003e\n\n## Acknowledgements\n\nThis study is supported by NTU NAP, MOE AcRF Tier 2 (T2EP20221-0033), and under the RIE2020 Industry Alignment Fund – Industry Collaboration Projects (IAF-ICP) Funding Initiative, as well as cash and in-kind contribution from the industry partner(s).\n\nWe thank the following repositories for their contributions in our implementation: [NeuS](https://github.com/Totoro97/NeuS), [smplx](https://github.com/vchoutas/smplx), [vposer](https://github.com/nghorbani/human_body_prior), [Smplx2FBX](https://github.com/mrhaiyiwang/Smplx2FBX).\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fhongfz16%2FAvatarCLIP","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fhongfz16%2FAvatarCLIP","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fhongfz16%2FAvatarCLIP/lists"}