{"id":21976911,"url":"https://github.com/yochengliu/scasnet","last_synced_at":"2026-02-20T23:39:08.520Z","repository":{"id":117937098,"uuid":"88409450","full_name":"Yochengliu/ScasNet","owner":"Yochengliu","description":"Semantic Labeling in VHR Images via A Self-Cascaded CNN (ISPRS JPRS, IF=6.942)","archived":false,"fork":false,"pushed_at":"2021-09-30T11:52:30.000Z","size":564,"stargazers_count":25,"open_issues_count":0,"forks_count":6,"subscribers_count":4,"default_branch":"master","last_synced_at":"2025-01-28T03:41:25.415Z","etag":null,"topics":["computer-vision","context-aware","deep-learning","multi-context","multi-scale","remote-sensing","semantic-labelling","semantic-segmentation"],"latest_commit_sha":null,"homepage":"https://www.sciencedirect.com/science/article/pii/S0924271617303854","language":null,"has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/Yochengliu.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null}},"created_at":"2017-04-16T11:39:53.000Z","updated_at":"2023-06-26T12:28:02.000Z","dependencies_parsed_at":"2023-09-24T16:58:15.880Z","dependency_job_id":null,"html_url":"https://github.com/Yochengliu/ScasNet","commit_stats":{"total_commits":37,"total_committers":1,"mean_commits":37.0,"dds":0.0,"last_synced_commit":"0f70a661dabead3f0551d1d3db91e0312d2665eb"},"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Yochengliu%2FScasNet","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Yochengliu%2FScasNet/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Yochengliu%2FScasNet/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Yochengliu%2FScasNet/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/Yochengliu","download_url":"https://codeload.github.com/Yochengliu/ScasNet/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":245037958,"owners_count":20550958,"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":["computer-vision","context-aware","deep-learning","multi-context","multi-scale","remote-sensing","semantic-labelling","semantic-segmentation"],"created_at":"2024-11-29T16:12:40.416Z","updated_at":"2025-10-24T06:04:06.801Z","avatar_url":"https://github.com/Yochengliu.png","language":null,"funding_links":[],"categories":[],"sub_categories":[],"readme":"Semantic Labeling in VHR Images via A Self-Cascaded CNN (ScasNet)\n===\nby [Yongcheng Liu](https://yochengliu.github.io/), [Bin Fan*](http://www.nlpr.ia.ac.cn/fanbin/), [Lingfeng Wang](https://scholar.google.com/citations?user=PaRak2AAAAAJ\u0026hl=en), [Jun Bai](https://www.researchgate.net/profile/Jun_Bai), [Shiming Xiang](https://scholar.google.com/citations?user=0ggsACEAAAAJ\u0026hl=zh-CN), [Chunhong Pan](https://www.researchgate.net/lab/Chunhong-Pan-Lab).  \n\n[vai]: ./images/vai.jpg\n![vai]\n\n## ScasNet\n\n#### VGG ScasNet\n\n- The encoder is based on VGG-Net variant (Chen et al., 2015), which is to obtain finer feature maps (about 1/8 of input size rather than 1/32).\n- On the last layer of encoder, multi-scale contexts are captured by dilated convolutional operations with dilation rates of 24, 18, 12, 6.\n- As a trade-off, we only choose three shallow layers for refinement. Moreover, BN layer is not used in VGG ScasNet.    \n\n#### ResNet ScasNet\n\nThe configuration of ResNet ScasNet is almost the same as VGG ScasNet, except for four aspects: \n\n- the encoder is based on ResNet variant (Zhao et al., 2016) \n- four shallow layers are used for refinement \n- seven residual correction schemes are designed for feature fusions\n- BN layer is used.  \n\n## Finetuning\n\n#### For initializing the encoder part in ScasNet\n    \n- The encoder in VGG ScasNet is finetuned with [VGG-Net_variant_caffemodel](http://liangchiehchen.com/projects/DeepLabv2_vgg.html)\n   \n- The encoder in ResNet ScasNet is finetuned with [ResNet_variant_caffemodel](https://drive.google.com/open?id=0BzaU285cX7TCNVhETE5vVUdMYk0)  \n\n## Caffe\n\n- The Caffe we used to train VGG ScasNet is released on [DeepLab_v2](https://bitbucket.org/aquariusjay/deeplab-public-ver2).\n   \n- The Caffe we used to train ResNet\tScasNet is released on [PSPNet](https://github.com/hszhao/PSPNet).\n      \n#### Installation\n\nPlease follow the instructions of [Caffe](https://github.com/BVLC/caffe), [DeepLab_v2](https://bitbucket.org/aquariusjay/deeplab-public-ver2) and [PSPNet](https://github.com/hszhao/PSPNet).  \n\nThe code has been tested successfully on Ubuntu 14.04 with CUDA 8.0.    \n\n## References\n1. Chen, L., Papandreou, G., Kokkinos, I., Murphy, K., Yuille, A. L., 2015. Semantic image segmentation with deep convolutional nets and fully connected crfs. In: International Conference on Learning Representations.   \n2. Zhao, H., Shi, J., Qi, X., Wang, X., Jia, J., 2016. Pyramid scene parsing network. arXiv preprint arXiv:1612.01105.\n\n## Citation\n\nWe would be very glad if ScasNet is helpful for your research, and please consider citing our paper ([arXiv](https://arxiv.org/abs/1807.11236)):   \n\n        @article{liu2018scasnet,   \n          author = {Yongcheng Liu and    \n                    Bin Fan and    \n                    Lingfeng Wang and   \n                    Jun Bai and   \n                    Shiming Xiang and   \n                    Chunhong Pan},   \n          title = {Semantic Labeling in Very High Resolution Images via A Self-Cascaded Convolutional Neural Network},   \n          journal = {ISPRS J. Photogram. and Remote Sensing.},   \n          volume = {145},  \n          pages = {78--95},  \n          year = {2018}   \n        }   \n\n## Contact\n\nWe would be very glad if you have some ideas or questions about ScasNet to share with us, please contact \u003cyongcheng.liu@nlpr.ia.ac.cn\u003e\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fyochengliu%2Fscasnet","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fyochengliu%2Fscasnet","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fyochengliu%2Fscasnet/lists"}