{"id":16604741,"url":"https://github.com/dmmaze/unet3plus-pytorch","last_synced_at":"2025-10-29T15:31:40.815Z","repository":{"id":116541649,"uuid":"489684111","full_name":"dmMaze/UNet3Plus-pytorch","owner":"dmMaze","description":"A pytorch implementation of UNetV3Plus","archived":false,"fork":false,"pushed_at":"2022-11-21T02:05:16.000Z","size":389,"stargazers_count":23,"open_issues_count":1,"forks_count":8,"subscribers_count":1,"default_branch":"master","last_synced_at":"2025-04-09T08:44:19.760Z","etag":null,"topics":["deep-learning","pytorch-implementation","segmentation","unet","unet-image-segmentation","unet3plus"],"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/dmMaze.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":"2022-05-07T13:41:27.000Z","updated_at":"2025-03-26T03:50:09.000Z","dependencies_parsed_at":null,"dependency_job_id":"f9c0020e-3f74-4fce-99c1-739b617c025d","html_url":"https://github.com/dmMaze/UNet3Plus-pytorch","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/dmMaze/UNet3Plus-pytorch","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/dmMaze%2FUNet3Plus-pytorch","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/dmMaze%2FUNet3Plus-pytorch/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/dmMaze%2FUNet3Plus-pytorch/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/dmMaze%2FUNet3Plus-pytorch/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/dmMaze","download_url":"https://codeload.github.com/dmMaze/UNet3Plus-pytorch/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/dmMaze%2FUNet3Plus-pytorch/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":281646120,"owners_count":26537234,"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","status":"online","status_checked_at":"2025-10-29T02:00:06.901Z","response_time":59,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"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":["deep-learning","pytorch-implementation","segmentation","unet","unet-image-segmentation","unet3plus"],"created_at":"2024-10-12T00:59:11.000Z","updated_at":"2025-10-29T15:31:40.810Z","avatar_url":"https://github.com/dmMaze.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"\n\nA pytorch implementation of UNetV3Plus.  \n\n![](assets/arch.png)\n\nUNetV3Plus is originally designed for medical image segmentation, I modify it to use custom encoder such as resnet and support multi-label segmentation, here is the result on Pascal VOC2012:\n\nTraining: 512x512 random crop  \nvalidation: 512x512 center crop\n\n|  Model          | Batch Size  |  mIoU        |\n| :--------        | :-------------: | :----:   | \n| UNetV3Plus-ResNet34       | 16*4      |   0.739     |\n\n  \nDownload from [Google Drive](https://drive.google.com/drive/folders/17fRLtYMA2LZdL7QDr7OF76uhN9GIK5uP?usp=sharing)\n\n16*4 means batch size 16 and 4 gradient accumulation steps.  \nRun ```python train.py --cfg config/resnet34_voc.yaml``` to reproduce the result.   \n\nAlthough Multi-label MS-SSIM loss is implemented but not used in the training, modify the 'loss_type' in the config file to 'u3p' to use it.   \nThe config file of the original model in the official paper is config/original_voc.yaml.   \nPlease reference config/config.py for more info about model arch or training settings. Custom Dataset is not supported yet.  \n\nUse tensorboard or wandb to log training metrics.  \n![](assets/logging.png)\n\n\n# Prepare Datasets\n## Pascal VOC\nDownload [VOC2012](http://host.robots.ox.ac.uk/pascal/VOC/voc2012/VOCtrainval_11-May-2012.tar) and [trainaug](https://www.dropbox.com/s/oeu149j8qtbs1x0/SegmentationClassAug.zip?dl=0), \nExtract trainaug labels (SegmentationClassAug) to the VOC2012 directory.  \nMore info about trainaug can be found in [DeepLabV3Plus](https://github.com/VainF/DeepLabV3Plus-Pytorch/blob/master/README.md).  \n\n```\n/data\n    /VOCdevkit  \n        /VOC2012\n            /SegmentationClass\n            /SegmentationClassAug  # \u003c= the trainaug labels\n                2007_000032.png\n                ...\n            /JPEGImages\n            ...\n        ...\n    /VOCtrainval_11-May-2012.tar\n    ...\n```\n\n# Acknowledgement\n* [UNet 3+: A Full-Scale Connected UNet for Medical Image Segmentation](https://arxiv.org/abs/2004.08790)\n* VOC2012 data-pipeline and eval-metrics are modified from  [DeepLabV3Plus](https://github.com/VainF/DeepLabV3Plus-Pytorch)\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fdmmaze%2Funet3plus-pytorch","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fdmmaze%2Funet3plus-pytorch","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fdmmaze%2Funet3plus-pytorch/lists"}