{"id":21205715,"url":"https://github.com/yilingqiao/dmrf","last_synced_at":"2025-04-09T23:18:57.540Z","repository":{"id":188590400,"uuid":"679032386","full_name":"YilingQiao/DMRF","owner":"YilingQiao","description":"Dynamic Mesh-Aware Radiance Fields (ICCV2023): Raytracing rendering and interactive simulating mesh with NeRF","archived":false,"fork":false,"pushed_at":"2024-02-28T19:59:41.000Z","size":193651,"stargazers_count":176,"open_issues_count":2,"forks_count":8,"subscribers_count":2,"default_branch":"cleaning","last_synced_at":"2025-04-09T23:18:29.212Z","etag":null,"topics":["cuda","nerf","raytracing","simulation"],"latest_commit_sha":null,"homepage":"https://mesh-aware-rf.github.io/","language":"Jupyter Notebook","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/YilingQiao.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE.txt","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":"2023-08-16T00:38:02.000Z","updated_at":"2025-03-04T08:59:16.000Z","dependencies_parsed_at":"2024-01-09T18:16:15.575Z","dependency_job_id":null,"html_url":"https://github.com/YilingQiao/DMRF","commit_stats":{"total_commits":437,"total_committers":40,"mean_commits":10.925,"dds":0.3661327231121282,"last_synced_commit":"0bb0bb5b1eeb4fab6bed66851b3bf8b8085f0915"},"previous_names":["yilingqiao/dmrf"],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/YilingQiao%2FDMRF","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/YilingQiao%2FDMRF/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/YilingQiao%2FDMRF/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/YilingQiao%2FDMRF/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/YilingQiao","download_url":"https://codeload.github.com/YilingQiao/DMRF/tar.gz/refs/heads/cleaning","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":248125639,"owners_count":21051776,"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":["cuda","nerf","raytracing","simulation"],"created_at":"2024-11-20T20:46:41.687Z","updated_at":"2025-04-09T23:18:57.518Z","avatar_url":"https://github.com/YilingQiao.png","language":"Jupyter Notebook","funding_links":[],"categories":[],"sub_categories":[],"readme":"\n\n\n# Dynamic Mesh-Aware Radiance Fields\n\n[Yi-Ling Qiao*](https://ylqiao.net/), [Alexander Gao*](https://gaoalexander.github.io/), [Yiran Xu](https://twizwei.github.io/), [Yue Feng](https://yuefeng21.github.io/), [Jia-Bin Huang](https://jbhuang0604.github.io/), [Ming C. Lin](https://www.cs.umd.edu/~lin/)\n\n[[Project]](https://mesh-aware-rf.github.io/) [[Paper]](https://arxiv.org/abs/2309.04581)  [[Video]](https://youtu.be/v9k_UzUKtcE) \n\n\n\n## Introduction\nThis is an implementation of the ICCV 2023 paper Dynamic Mesh-Aware Radiance Fields.\nThe NeRF volume rendering is largely based on NVIDA's great work [InstantNGP](https://github.com/NVlabs/instant-ngp). The physics part also runs with NVIDIA's [Warp](https://github.com/NVIDIA/warp) and/or DeepMind's [Mujoco](https://github.com/deepmind/mujoco). A more detailed tutorial for ray tracing can be found in [Ray Tracing in One Weekend](https://github.com/RayTracing/raytracing.github.io).\n\n\n## Structure\nWe have now uploaded all core C++/CUDA files but are still cleaning the python scripts, config files, and datasets for running each experiment.\n\nThe entry to the main rendering loop for raytracing-NeRF coupling is the `render_nerf_rt` function at `./src/testbed_nerf.cu`. \n\nThe folder `./simpleRt/` also has CUDA-based raytracing-related functions. \n\nPhysics simulation module will be called in python scripts.\n\n## Setup\nThis project has been tested with Ubuntu 20.04, gcc 9.4.0, CUDA 11.8.\n```bash\ngit clone git@github.com:YilingQiao/DMRF.git\ncd DMRF\ngit submodule update --init --recursive\nmkdir build\ncd build\ncmake ..\nmake -j\ncd ..\n```\n\nIf you have any issues building/running this project, please try to make sure you can run [InstantNGP](https://github.com/NVlabs/instant-ngp) first (especially its commit `c4d622e`, which is a start point of our project).\n\n## TODOs\n1. Other scenes presented in the paper.\n\n## Demos\n\nLet's download the data and pretrained NeRF from this [google drive](https://drive.google.com/drive/folders/1n8BJhkSCBqXTN-mkdMdfrM5IpsMzkj61?usp=drive_link).\n\nPut them under `DMRF/extra_data`.\n\n### Reflective ball\nWe can insert a reflective ball into the `nerf360/garden` scene. The ball will also cast shdows onto the ground. \n\nScene desctiption of the ball, lightsource, and shadow mesh can be found in the configuration file `./scripts/exp/garden_ball.json\"`.\n\n```\npython ./scripts/exp_garden_ball.py --mode nerf --load_snapshot ./extra_data/nerf360/garden/35000.msgpack  --width 800 --height 600 --gui\n```\n\n\u003cdiv align=\"center\"\u003e\n\u003cimg width=\"400px\" src=\"https://github.com/YilingQiao/linkfiles/raw/master/23DMRF/garden_ball_shadow.gif\"\u003e \n\u003c/div\u003e\n\n\n### Infinite mirror room\nLet's go one step further and add more objects into the scene.\n\nWith relfective walls around, now the ray can bounce multiple times inside a mirror room.\n\nSince the scene is complex, we choose not to render it in GUI. Images will be saved to `./render_output/garden_mirror/`.\n\n```\npython ./scripts/exp_garden_mirror.py --mode nerf --load_snapshot ./extra_data/nerf360/garden/35000.msgpack --width 200 --height 150 --video_spp 4 \n```\n\n\u003cdiv align=\"center\"\u003e\n\u003cimg width=\"400px\" src=\"https://github.com/YilingQiao/linkfiles/raw/master/23DMRF/garden_mirror_monkey.gif\"\u003e \n\u003c/div\u003e\n\n### Interactive game\nWe can also add a controllable object into NeRF and see how it interact with the scene.\n\nHere we use [Warp](https://github.com/NVIDIA/warp) to compute the dynamics. You might need to install this package first.\n\nUse `W/S/A/D/Z/X` to control the ball's moving direction. Use `space` to stop the ball\n\n```\npython ./scripts/exp_game.py --mode nerf --load_snapshot ./extra_data/nerf360/kitchen/35000.msgpack --width 800 --height 600 --gui --simulation \n```\n\n\u003cdiv align=\"center\"\u003e\n\u003cimg width=\"400px\" src=\"https://github.com/YilingQiao/linkfiles/raw/master/23DMRF/game_counter.gif\"\u003e \n\u003c/div\u003e\n\n## Run with your own data\n\n### Preprocess\nYou can refer to this [tutorial](preprocess/README.md) for preprocessing your own data.\n\n### Config file\nScene configuration files are stored in `scripts/exp/*.json`.\n\nGeometries in the scene can be provided to the scene by setting `spheres` and `objfile`. You can set the sphere's `center` and `radius`.\n\nRendering materials of the geometries can be set using `materials` entries. Available types include `dielectric`, `lambertian`, `metal`, `shadow`, `lightsource`. Note that the last two are used for cast shadows and not actually visible in the scene.\n\nThe `yiling/231018_lightsrc` branch has more experimental features for controlling the light source, and it can render faster for lambertian materials.\n\nNote that the material `id` in `geometries` and `materials` are different by 1, which might be a little confusing. (i.e. material id in geometries =  `id` in `mateirals` + 1)\n\n\n\n## BibTex\n```\n@article{qiao2023dmrf,\n  author    = {Yi-Ling Qiao and Alexander Gao and Yiran Xu and Yue Feng and Jia-Bin Huang and Ming C. Lin},\n  title     = {Dynamic Mesh-Aware Radiance Fields},\n  journal   = {ICCV},\n  year      = {2023},\n}\n```\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fyilingqiao%2Fdmrf","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fyilingqiao%2Fdmrf","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fyilingqiao%2Fdmrf/lists"}