{"id":17991580,"url":"https://github.com/speedinghzl/dsrg","last_synced_at":"2025-03-25T23:32:21.214Z","repository":{"id":59867024,"uuid":"125361591","full_name":"speedinghzl/DSRG","owner":"speedinghzl","description":"Weakly-Supervised Semantic Segmentation Network with Deep Seeded Region Growing (CVPR 2018).","archived":false,"fork":false,"pushed_at":"2018-12-30T10:50:38.000Z","size":233,"stargazers_count":250,"open_issues_count":15,"forks_count":34,"subscribers_count":11,"default_branch":"master","last_synced_at":"2025-03-20T23:01:38.359Z","etag":null,"topics":["caffe","pascal-voc","segmentation-network","semantic-segmentation"],"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/speedinghzl.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":"2018-03-15T12:06:52.000Z","updated_at":"2024-10-20T02:30:49.000Z","dependencies_parsed_at":"2022-09-23T22:50:24.391Z","dependency_job_id":null,"html_url":"https://github.com/speedinghzl/DSRG","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/speedinghzl%2FDSRG","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/speedinghzl%2FDSRG/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/speedinghzl%2FDSRG/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/speedinghzl%2FDSRG/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/speedinghzl","download_url":"https://codeload.github.com/speedinghzl/DSRG/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":245562241,"owners_count":20635896,"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":["caffe","pascal-voc","segmentation-network","semantic-segmentation"],"created_at":"2024-10-29T19:22:51.879Z","updated_at":"2025-03-25T23:32:17.133Z","avatar_url":"https://github.com/speedinghzl.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Weakly-Supervised Semantic Segmentation Network with Deep Seeded Region Growing (CVPR2018)\nBy [Zilong Huang](http://speedinghzl.github.io), [Xinggang Wang](http://mclab.eic.hust.edu.cn/~xwang/index.htm), [Jiasi Wang](https://github.com/JiasiWang), [Wenyu Liu](http://mclab.eic.hust.edu.cn/MCWebDisplay/PersonDetails.aspx?Name=Wenyu%20Liu) and [Jingdong Wang](https://jingdongwang2017.github.io/).\n\nThis code is a implementation of the weakly-supervised semantic segmentation experiments in the paper [DSRG](http://openaccess.thecvf.com/content_cvpr_2018/papers/Huang_Weakly-Supervised_Semantic_Segmentation_CVPR_2018_paper.pdf). The code is developed based on the Caffe framework.\n\n## Introduction\n![Overview of DSRG](https://user-images.githubusercontent.com/4509744/50546511-5bb19f00-0bee-11e9-85de-4660369dbb59.png)\nOverview of the proposed approach. The Deep Seeded Region Growing module takes the seed cues and segmentation map as input to produces latent pixel-wise supervision which is more accurate and more complete than seed cues. Our method iterates between reļ¬ning pixel-wise supervision and optimizing the parameters of a segmentation network.\n\n\n### License\n\nDSRG is released under the MIT License (refer to the LICENSE file for details).\n\n### Citing DSRG\n\nIf you find DSRG useful in your research, please consider citing:\n\n @inproceedings{huang2018dsrg,\n title={Weakly-Supervised Semantic Segmentation Network with Deep Seeded Region Growing},\n author={Huang, Zilong and Wang, Xinggang and Wang, Jiasi and Liu, Wenyu and Wang, Jingdong},\n booktitle={Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition},\n pages={7014--7023},\n year={2018}\n }\n \n## Installing dependencies\n\n* Python packages:\n```bash\n $ pip install -r python-dependencies.txt\n```\n* **caffe (deeplabv2 version)**: deeplabv2 caffe installation instructions are available at `https://bitbucket.org/aquariusjay/deeplab-public-ver2`. Note, you need to compile **caffe** with python wrapper and support for python layers. Then add the caffe python path into [training/tools/findcaffe.py](https://github.com/speedinghzl/DSRG/blob/master/training/tools/findcaffe.py#L21).\n\n* Fully connected CRF wrapper (requires the **Eigen3** package).\n```bash\n $ pip install CRF/\n```\n\n## Training the DSRG model\n\n* Go into the training directory: \n\n```bash\n $ cd training\n $ mkdir localization_cues\n```\n\n* Download the initial [VGG16](https://drive.google.com/open?id=1nq49w4os6BZ1JcrM4xqZKZh1wR3-32wi) model pretrained on Imagenet and put it in *training/* folder.\n\n* Download CAM [seed](https://drive.google.com/open?id=1cHyhjul9srPlwcl4xqrYR9MwzhFGwKXU) and put it in *training/localization_cues* folder. We use [CAM](http://cnnlocalization.csail.mit.edu/) for localizing the foreground seed classes and utilize the saliency detection technology [DRFI](http://supermoe.cs.umass.edu/~hzjiang/drfi/) for localizing background seed. We provide the python interface to DRFI [here](https://github.com/speedinghzl/drfi_cpp) for convenience if you want to generate the seed by yourself.\n\n```bash\n $ cd training/experiment/seed_mc\n $ mkdir models\n```\n* Set *root_folder* parameter in **train-s.prototxt, train-f.prototxt** and *PASCAL_DIR* in **run-s.sh** to the directory with **PASCAL VOC 2012** images\n\n* Run:\n\n```bash\n $ bash run.sh\n```\n The trained model will be created in `models`\n \n \n## Acknowledgment\nThis code is heavily borrowed from [SEC](https://github.com/kolesman/SEC).\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fspeedinghzl%2Fdsrg","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fspeedinghzl%2Fdsrg","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fspeedinghzl%2Fdsrg/lists"}