{"id":22324054,"url":"https://github.com/lzhengning/subdivnet","last_synced_at":"2025-05-16T18:09:01.570Z","repository":{"id":41460543,"uuid":"374541355","full_name":"lzhengning/SubdivNet","owner":"lzhengning","description":"Subdivision-based Mesh Convolutional Networks. ","archived":false,"fork":false,"pushed_at":"2024-10-21T02:24:55.000Z","size":786,"stargazers_count":269,"open_issues_count":12,"forks_count":35,"subscribers_count":6,"default_branch":"master","last_synced_at":"2025-04-03T19:13:41.688Z","etag":null,"topics":["convolutional-neural-networks","jittor","mesh-networks","subdivision"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/lzhengning.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":"2021-06-07T05:08:56.000Z","updated_at":"2025-03-28T11:04:28.000Z","dependencies_parsed_at":"2024-01-18T15:05:30.737Z","dependency_job_id":"4d5e55e8-ebf7-48ea-a6e5-a6971c378f53","html_url":"https://github.com/lzhengning/SubdivNet","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/lzhengning%2FSubdivNet","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/lzhengning%2FSubdivNet/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/lzhengning%2FSubdivNet/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/lzhengning%2FSubdivNet/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/lzhengning","download_url":"https://codeload.github.com/lzhengning/SubdivNet/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":248606860,"owners_count":21132428,"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":["convolutional-neural-networks","jittor","mesh-networks","subdivision"],"created_at":"2024-12-04T02:00:18.905Z","updated_at":"2025-04-12T17:38:10.067Z","avatar_url":"https://github.com/lzhengning.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Subdivision-based Mesh Convolutional Networks\n\nThe implementation of `SubdivNet` in our paper, [Subdivion-based Mesh Convolutional Networks](https://cg.cs.tsinghua.edu.cn/papers/TOG-2022-SubdivNet.pdf)\n\n![teaser](teaser.jpg)\n\n## News\n* 🔥This paper was accepted by [ACM TOG](https://dl.acm.org/doi/10.1145/3506694). \n\n## Features\n* Provides implementations of mesh classification and segmentation on various datasets.\n* Provides ready-to-use datasets, pretrained models, training and evaluation scripts.\n* Supports a batch of meshes with different number of faces.\n\n## Requirements\n* python3.7+\n* CUDA 10.1+\n* [Jittor](https://github.com/Jittor/jittor)\n\nTo install other python requirements:\n\n```\npip install -r requirements.txt\n```\n\n## Fetch Data\nThis repo provides training scripts for classification and segementation, \non the following datasets,\n\n- shrec11-split10\n- shrec11-split16\n- cubes\n- manifold40 (based on ModelNet40)\n- humanbody\n- coseg-aliens\n- coseg-vases\n\nTo download the preprocessed data, run\n\n```\nsh scripts/\u003cDATASET_NAME\u003e/get_data.sh\n```\n\n\u003e The `Manfold40` dataset (before remeshed, without subdivision connectivity) can be downloaded via [this link](https://cg.cs.tsinghua.edu.cn/dataset/subdivnet/datasets/Manifold40.zip). \n\u003e Note that this version cannot be used as inputs of SubdivNet. To train SubdivNet, run scripts/manifold40/get_data.sh.\n\n## Training\nTo train the model(s) in the paper, run this command:\n\n```\nsh scripts/\u003cDATASET_NAME\u003e/train.sh\n```\n\nTo speed up training, you can use multiple gpus. First install `OpenMPI`: \n\n```\nsudo apt install openmpi-bin openmpi-common libopenmpi-dev\n```\n\nThen run the following command,\n\n```\nCUDA_VISIBLE_DEVICES=\"2,3\" mpirun -np 2 sh scripts/\u003cDATASET_NAME\u003e/train.sh\n```\n\n## Evaluation\n\nTo evaluate the model on a dataset, run:\n\n```\nsh scripts/\u003cDATASET_NAME\u003e/test.sh\n```\n\nThe pretrained weights are provided. Run the following command to download them.\n\n```\nsh scripts/\u003cDATASET_NAME\u003e/get_pretrained.sh\n```\n\n## Visualize\nAfter testing the segmentation network, there will be colored shapes in a `results` directory.\n\n## How to apply SubdivNet to your own data\nSubdivNet cannot be directly applied to any common meshes, because it requires the input to hold the subdivision connectivity.\n\nTo create your own data with subdivision connectivity, you may use the provided\ntool that implements the MAPS algorithm. You may also refer to [NeuralSubdivision](https://github.com/HTDerekLiu/neuralSubdiv), as they provide a MATLAB script for remeshing.\n\nTo run our implemented MAPS algorithm, first install the following python dependecies,\n\n```\ntriangle\npymeshlab\nshapely\nsortedcollections\nnetworkx\nrtree\n```\n\nThen see `datagen_maps.py` and modify the configurations to remesh your 3D shapes for subdivision connectivity.\n\n## Cite\nPlease cite our paper if you use this code in your own work:\n\n```\n@article{subdivnet-tog-2022,\n  author    = {Shi{-}Min Hu and\n               Zheng{-}Ning Liu and\n               Meng{-}Hao Guo and\n               Junxiong Cai and\n               Jiahui Huang and\n               Tai{-}Jiang Mu and\n               Ralph R. Martin},\n  title     = {Subdivision-based Mesh Convolution Networks},\n  journal   = {{ACM} Trans. Graph.},\n  volume    = {41},\n  number    = {3},\n  pages     = {25:1--25:16},\n  year      = {2022},\n  url       = {https://doi.org/10.1145/3506694}\n}\n```\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Flzhengning%2Fsubdivnet","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Flzhengning%2Fsubdivnet","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Flzhengning%2Fsubdivnet/lists"}