{"id":22704387,"url":"https://github.com/justanhduc/graphx-conv","last_synced_at":"2025-07-04T02:34:10.985Z","repository":{"id":58560779,"uuid":"177146919","full_name":"justanhduc/graphx-conv","owner":"justanhduc","description":"Official implementation of GraphX-Convolution","archived":false,"fork":false,"pushed_at":"2020-09-07T08:58:15.000Z","size":16159,"stargazers_count":62,"open_issues_count":6,"forks_count":17,"subscribers_count":5,"default_branch":"master","last_synced_at":"2025-07-02T11:57:52.882Z","etag":null,"topics":["3d-reconstruction","deep-learning","graphx-convolution","pytorch","shapenet"],"latest_commit_sha":null,"homepage":"https://justanhduc.github.io/2019/09/29/GraphX-Convolution.html","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/justanhduc.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}},"created_at":"2019-03-22T13:40:34.000Z","updated_at":"2024-12-19T07:21:58.000Z","dependencies_parsed_at":"2022-09-21T00:00:28.957Z","dependency_job_id":null,"html_url":"https://github.com/justanhduc/graphx-conv","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/justanhduc/graphx-conv","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/justanhduc%2Fgraphx-conv","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/justanhduc%2Fgraphx-conv/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/justanhduc%2Fgraphx-conv/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/justanhduc%2Fgraphx-conv/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/justanhduc","download_url":"https://codeload.github.com/justanhduc/graphx-conv/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/justanhduc%2Fgraphx-conv/sbom","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":263434321,"owners_count":23465937,"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-reconstruction","deep-learning","graphx-convolution","pytorch","shapenet"],"created_at":"2024-12-10T08:15:28.232Z","updated_at":"2025-07-04T02:34:10.964Z","avatar_url":"https://github.com/justanhduc.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# GraphX-Convolution\n\nThe official implementation of the ICCV 2019 paper \n\"GraphX-convolution for point cloud deformation in 2D-to-3D conversion\".\n\n![framework](./imgs/framework.jpg)\n\n__Update 1:__ We have achieved a better performance with `PointCloudResLowRankGraphXUpDecoder`, \nwhich is an upgraded version of the best deformation network in the paper. The new network works \ncomparably with the best network in the paper but consumes much less memory.\n\n__Update 2 (Sep 7, 2020):__ Refactor the code to work with the latest [Neuralnet-pytorch](https://github.com/justanhduc/neuralnet-pytorch).\n\n## Prerequisite\n\n[Pytorch](https://pytorch.org/get-started/locally/) (\u003e=1.0.0)\n\n[Neuralnet-pytorch](https://github.com/justanhduc/neuralnet-pytorch) (``pip install \"neuralnet-pytorch[gin] @ git+git://github.com/justanhduc/neuralnet-pytorch.git@6bda19fdc57f176cb82f58d287602f4ccf4cfc23\" --global-option=\"--cuda-ext\"``)\n\n[Gin-config](https://github.com/google/gin-config)\n\n[Matplotlib](https://matplotlib.org/)\n\n[TensorboardX](https://github.com/lanpa/tensorboardX)\n\nThe code has been tested on a Linux Ubuntu 16.04 machine.\n\n## Data\n\nThe data we used in our experiment provided by [ShapeNet](https://www.shapenet.org/). \nHowever, for convenience, we used the pre-processed data provided by Pixel2mesh. \nThe data can be downloaded from [here](https://github.com/nywang16/Pixel2Mesh).\n\nThe train/test split used in the paper is the default split provided by ShapeNet. \nFor more convenience, we provide the pre-processed train/test file lists in [data](./data).\n\nTo process the data into the format used by the code, execute \n\n```\ncd data\npython split.py path/to/the/downloaded/data/folder\n```\n\nA pre-processed zip file can be found [here](https://drive.google.com/file/d/1EzGkZ4XBhVNGZPnMXPSbkOaNz2sechbZ/view)\n\n## Training and testing\n\n### Training\n\nAfter the database is setup, we are ready to train the model. \nIn the [configs](./src/configs) folder, we have prepared the basic training configs for several models\nused in the paper.\nAfter the data is downloaded, specify the data root in these config files.\n\nBy default, the script creates a `results` folder containing checkpoint folders \ncorresponding to the model name.\nThe model name can be specified in the config files.\n\nThen, simply execute\n\n```\ncd src\npython train.py path/to/a/config/file (--gpu 0)\n```\n\nTo resume a half-done training, simply specify the checkpoint folder in the config file.\nThen run the training script as in the training step.\n\n### Evaluation\n\nBe sure to use the same config file and specify the checkpoint folder in the config file.\nAfter the model is fully trained, to test the model, use\n\n```\npython test.py path/to/a/config/file (--gpu 0)\n```\n\nThe script calculates the Chamfer distance (CD) scores similar to \n[Pixel2mesh](https://github.com/nywang16/Pixel2Mesh).\n\n## Results\n\n### Visualizations\n\n\u003cp align='center'\u003e\n  \u003cimg src='imgs/results/airplane.gif' width=\"250px\"\u003e\n  \u003cimg src='imgs/results/bench.gif' width=\"250px\"\u003e\n  \u003cimg src='imgs/results/car.gif' width=\"250px\"\u003e\n\u003c/p\u003e\n\n\u003cp align='center'\u003e\n  \u003cimg src='imgs/results/lamp.gif' width=\"250px\"\u003e\n  \u003cimg src='imgs/results/monitor.gif' width=\"250px\"\u003e\n  \u003cimg src='imgs/results/table.gif' width=\"250px\"\u003e\n\u003c/p\u003e\n\n### Numerical results\n\n|Methods|Chamfer|IoU|\n|:-------:|:---:|:-------:|\n|[3D-R2N2](http://3d-r2n2.stanford.edu/)|1.445|0.631|\n|[PSG](https://arxiv.org/abs/1612.00603)|0.593|0.640|\n|[Pixel2mesh](https://github.com/nywang16/Pixel2Mesh)|0.591|-|\n|[GAL](https://link.springer.com/chapter/10.1007/978-3-030-01237-3_49)|-|0.712|\n|Ours (UpResGraphX)|__0.252__|__0.725__|\n\n## Citation\n\nTBA\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fjustanhduc%2Fgraphx-conv","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fjustanhduc%2Fgraphx-conv","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fjustanhduc%2Fgraphx-conv/lists"}