{"id":18322517,"url":"https://github.com/tencentarc/hosnerf","last_synced_at":"2025-04-05T23:31:06.514Z","repository":{"id":155922685,"uuid":"632771335","full_name":"TencentARC/HOSNeRF","owner":"TencentARC","description":"HOSNeRF: Dynamic Human-Object-Scene Neural Radiance Fields from a Single Video","archived":false,"fork":false,"pushed_at":"2023-12-12T15:35:34.000Z","size":24864,"stargazers_count":68,"open_issues_count":4,"forks_count":7,"subscribers_count":7,"default_branch":"main","last_synced_at":"2025-03-21T13:23:21.198Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":null,"language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"apache-2.0","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/TencentARC.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,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2023-04-26T05:17:05.000Z","updated_at":"2025-03-11T04:48:50.000Z","dependencies_parsed_at":"2024-11-05T18:44:40.007Z","dependency_job_id":null,"html_url":"https://github.com/TencentARC/HOSNeRF","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/TencentARC%2FHOSNeRF","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/TencentARC%2FHOSNeRF/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/TencentARC%2FHOSNeRF/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/TencentARC%2FHOSNeRF/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/TencentARC","download_url":"https://codeload.github.com/TencentARC/HOSNeRF/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":247415783,"owners_count":20935383,"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":[],"created_at":"2024-11-05T18:24:59.089Z","updated_at":"2025-04-05T23:31:01.505Z","avatar_url":"https://github.com/TencentARC.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# [ICCV2023] HOSNeRF: Dynamic Human-Object-Scene Neural Radiance Fields from a Single Video\n\n  \nThis is the official repository of **HOSNeRF** [Project page](https://showlab.github.io/HOSNeRF) | [arXiv](https://arxiv.org/abs/2304.12281) | [Video](https://www.youtube.com/watch?v=wS5k5nNkPi4)\n\n[Jia-Wei Liu](https://jia-wei-liu.github.io/), [Yan-Pei Cao](https://yanpei.me),  [Tianyuan Yang](https://scholar.google.com.hk/citations?user=s2q3_A4AAAAJ\u0026hl=zh-CN),  [Zhongcong Xu](https://scholar.google.com/citations?user=-4iADzMAAAAJ\u0026hl=en), [Jussi Keppo](https://www.jussikeppo.com/), [Ying Shan](https://scholar.google.com/citations?user=4oXBp9UAAAAJ\u0026hl=en), [Xiaohu Qie](https://scholar.google.com/citations?hl=en\u0026user=mk-F69UAAAAJ\u0026view_op=list_works\u0026sortby=pubdate), [Mike Zheng Shou](https://sites.google.com/view/showlab)\n\n\u003e **TL;DR:** A novel 360° free-viewpoint rendering method that reconstructs neural radiance fields for dynamic human-object-scene from a single monocular in-the-wild video.\n\n\u003cp align=\"center\"\u003e\n\u003cimg src=\"/assets/HOSNeRF.gif\" width=\"1080px\"/\u003e  \n\u003cbr\u003e\n\u003cem\u003eHOSNeRF can render 360° free-viewpoint videos from a single monocular in-the-wild video.\u003c/em\u003e\n\u003c/p\u003e\n\n\u003cp align=\"center\"\u003e\n\u003cimg src=\"/assets/HOSNeRF.png\" width=\"1080px\"/\u003e  \n\u003cbr\u003e\n\u003cem\u003eHOSNeRF Framework.\u003c/em\u003e\n\u003c/p\u003e\n\n## 📢 News\n\n - [2023.12.11] We release the HOSNeRF codebase!\n\n- [2023.08.16] We release the HOSNeRF dataset!\n\n- [2023.08.12] HOSNeRF got accepted by [**ICCV 2023**](https://iccv2023.thecvf.com/)!\n\n  \n\n- [2023.04.24] We release the arXiv paper!\n\n  \n\n  \n\n## 📝 Preparation\n\n  \n\n### Installation\n\n```\ngit clone https://github.com/TencentARC/HOSNeRF.git\ncd HOSNeRF\npip install -r requirements.txt\n```\n\n### Download SMPL model\n\nDownload the gender neutral SMPL model from [here](https://smplify.is.tue.mpg.de/), and unpack **mpips_smplify_public_v2.zip**.\n\nCopy the smpl model.\n\n    SMPL_DIR=/path/to/smpl\n    MODEL_DIR=$SMPL_DIR/smplify_public/code/models\n    cp $MODEL_DIR/basicModel_neutral_lbs_10_207_0_v1.0.0.pkl third_parties/smpl/models of 2nd_State_Conditional_Human-Object and 3rd_Complete_HOSNeRF\n\nFollow [this page](https://github.com/vchoutas/smplx/tree/master/tools) to remove Chumpy objects from the SMPL model.\n\n### HOSNeRF dataset\n\n\nWe release the HOSNeRF dataset on [link](https://drive.google.com/drive/folders/1viuXcihwFpLIjl6TmLyF5VARB7GxEfEv). HOSNeRF dataset consists of 6 real-world dynamic human-object-scene sequences: Backpack, Tennis, Suitcase, Playground, Dance, Lounge. Please run the optical flow estimation method using [RAFT](https://github.com/princeton-vl/RAFT) to get the optical flows of each scene. \n\n## 🏋️‍️ Experiment\n\n  \n\n### Training\n\n  \n\n**Stage 1: Train the state-conditional background model.**\n\n```bash\n$ cd 1st_State-Conditional_Scene\n$ CUDA_VISIBLE_DEVICES=0,1,2,3 python run.py --ginc configs/state_mipnerf360/Backpack.gin --scene Backpack --logbase 'path to logbase'\n```\n\n**Stage 2: Train the state-conditional dynamic human-object model.** \n\n(Please also change the datadir in configs/default.yaml and core/data/dataset_args.py)\n\n```bash\n$ cd 2nd_State_Conditional_Human-Object\n$ CUDA_VISIBLE_DEVICES=0,1,2,3 python run.py --ginc configs/human-object/Backpack.gin --scene Backpack --logbase 'path to logbase' --cfg configs/human_nerf/wild/monocular/adventure.yaml --seed 777\n```\n\n**Stage 3: Train the complete HOSNeRF model using the trained background and human-object checkpoints.** \n\n(Please also change the datadir in configs/default.yaml and core/data/dataset_args.py)\n\n```bash\n$ cd 3rd_Complete_HOSNeRF\n$ CUDA_VISIBLE_DEVICES=0,1,2,3 python run.py --ginc configs/HOSNeRF/Backpack.gin --scene Backpack --logbase 'path to logbase' --cfg configs/human_nerf/wild/monocular/adventure.yaml --seed 777\n```\n\n### Evaluation\n\nWe include the test codes in the model's test_step function. It will automatically run the test metrics (`PSNR`, `SSIM`, and `LPIPS`) for test images and all images after training.\n\n\n### Render 360° free-viewpoint videos\n\nPlease change the freeview index in the configs/default.yaml to render the free-viewpoint videos of that timestep. It will automatically render free-viewpoint videos (the model's test_step function) after training.\n\n\n### Render canonical human-object videos\n\nIt will automatically render the canonical human-object videos (the model's test_step function) after training.\n\n\n### Resume training\n\nTo resume training or resume testing after training, please add --resume_training True for each training script.\n\n\n### HOSNeRF checkpoints\n\nWe release the 6 HOSNeRF checkpoints on [link](https://drive.google.com/drive/folders/15I7z7qjBL6rQ3z91_rzhR284L1vfxxOX?usp=drive_link) for reference.\n\n\n## 🎓 Citation\n\nIf you find our work helps, please cite our paper.\n\n```bibtex\n@inproceedings{liu2023hosnerf,\n  title={Hosnerf: Dynamic human-object-scene neural radiance fields from a single video},\n  author={Liu, Jia-Wei and Cao, Yan-Pei and Yang, Tianyuan and Xu, Zhongcong and Keppo, Jussi and Shan, Ying and Qie, Xiaohu and Shou, Mike Zheng},\n  booktitle={Proceedings of the IEEE/CVF International Conference on Computer Vision},\n  pages={18483--18494},\n  year={2023}\n}\n```\n\n## ✉️ Contact\nThis repo is maintained by [Jiawei Liu](https://jia-wei-liu.github.io/). Questions and discussions are welcome via jiawei.liu@u.nus.edu.\n\n## 🙏 Acknowledgements\nThis codebase is based on [HumanNeRF](https://github.com/chungyiweng/humannerf) and [NeRF-Factory](https://github.com/kakaobrain/nerf-factory). The preprocessing code is based on [NeuMan](https://github.com/apple/ml-neuman). Thanks for open-sourcing!\n\n## LICENSE\nCopyright (c) 2023 Show Lab, National University of Singapore. All Rights Reserved. Licensed under the Apache License, Version 2.0 (see [LICENSE](https://github.com/TencentARC/HOSNeRF/blob/main/LICENSE) for details)","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Ftencentarc%2Fhosnerf","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Ftencentarc%2Fhosnerf","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Ftencentarc%2Fhosnerf/lists"}