{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/refinenet-multi-path-refinement-networks-for","title":"RefineNet: Multi-Path Refinement Networks for High-Resolution Semantic Segmentation","arxiv_id":"1611.06612","date":"2016-11-20","proceeding":"CVPR 2017 7","authors":["Guosheng Lin","Anton Milan","Chunhua Shen","Ian Reid"],"abstract":"Recently, very deep convolutional neural networks (CNNs) have shown\noutstanding performance in object recognition and have also been the first\nchoice for dense classification problems such as semantic segmentation.\nHowever, repeated subsampling operations like pooling or convolution striding\nin deep CNNs lead to a significant decrease in the initial image resolution.\nHere, we present RefineNet, a generic multi-path refinement network that\nexplicitly exploits all the information available along the down-sampling\nprocess to enable high-resolution prediction using long-range residual\nconnections. In this way, the deeper layers that capture high-level semantic\nfeatures can be directly refined using fine-grained features from earlier\nconvolutions. The individual components of RefineNet employ residual\nconnections following the identity mapping mindset, which allows for effective\nend-to-end training. Further, we introduce chained residual pooling, which\ncaptures rich background context in an efficient manner. We carry out\ncomprehensive experiments and set new state-of-the-art results on seven public\ndatasets. In particular, we achieve an intersection-over-union score of 83.4 on\nthe challenging PASCAL VOC 2012 dataset, which is the best reported result to\ndate.","url_abs":"http://arxiv.org/abs/1611.06612v3","url_pdf":"http://arxiv.org/pdf/1611.06612v3.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"refinenet-multi-path-refinement-networks-for","repo_url":"https://github.com/guosheng/refinenet","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"NOASSERTION"}},{"paper_slug":"refinenet-multi-path-refinement-networks-for","repo_url":"https://github.com/alililia/ascend_RefineNet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"mindspore","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"refinenet-multi-path-refinement-networks-for","repo_url":"https://github.com/hiyaroy12/multitask_learning","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null},{"paper_slug":"refinenet-multi-path-refinement-networks-for","repo_url":"https://github.com/hugozanini/realtime-semantic-segmentation","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"refinenet-multi-path-refinement-networks-for","repo_url":"https://github.com/kdha0727/cancer-instance-segmentation-from-tissue","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"refinenet-multi-path-refinement-networks-for","repo_url":"https://github.com/kingcong/RefineNet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"mindspore","reach":{"status":"gone","observed_at":"2026-09-18","how":"tree_404+repo_404"}},{"paper_slug":"refinenet-multi-path-refinement-networks-for","repo_url":"https://github.com/kingcong/gpu_RefineNet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"mindspore","reach":{"status":"gone","observed_at":"2026-09-18","how":"tree_404+repo_404"}},{"paper_slug":"refinenet-multi-path-refinement-networks-for","repo_url":"https://github.com/yangyucheng000/RefineNet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"mindspore","reach":{"status":"ok"}},{"paper_slug":"refinenet-multi-path-refinement-networks-for","repo_url":"https://github.com/2023-MindSpore-1/ms-code-216/tree/main/RefineNet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"mindspore","reach":null},{"paper_slug":"refinenet-multi-path-refinement-networks-for","repo_url":"https://github.com/2023-MindSpore-4/Code6/tree/main/RefineNet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"none","reach":null},{"paper_slug":"refinenet-multi-path-refinement-networks-for","repo_url":"https://github.com/MindSpore-paper-code-3/code5/tree/main/RefineNet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"mindspore","reach":null},{"paper_slug":"refinenet-multi-path-refinement-networks-for","repo_url":"https://github.com/code-implementation1/Code7/tree/main/RefineNet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"mindspore","reach":null},{"paper_slug":"refinenet-multi-path-refinement-networks-for","repo_url":"https://github.com/mindspore-ai/models/tree/master/research/cv/RefineNet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"mindspore","reach":null}],"tasks":[{"task_slug":"3d-absolute-human-pose-estimation","task_name":"3D Absolute Human Pose Estimation"},{"task_slug":"semantic-segmentation","task_name":"Semantic Segmentation"},{"task_slug":"high","task_name":"Vocal Bursts Intensity Prediction"}],"methods":[{"method_slug":"1x1-convolution","method_name":"1x1 Convolution"},{"method_slug":"average-pooling","method_name":"Average Pooling"},{"method_slug":"batch-normalization","method_name":"Batch Normalization"},{"method_slug":"bottleneck-residual-block","method_name":"Bottleneck Residual Block"},{"method_slug":"convolution","method_name":"Convolution"},{"method_slug":"global-average-pooling","method_name":"Global Average Pooling"},{"method_slug":"kaiming-initialization","method_name":"Kaiming Initialization"},{"method_slug":"max-pooling","method_name":"Max Pooling"},{"method_slug":"relu","method_name":"ReLU"},{"method_slug":"residual-block","method_name":"Residual Block"},{"method_slug":"residual-connection","method_name":"Residual Connection"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/semantic-segmentation-on-ade20k","task":"Semantic Segmentation","dataset":"ADE20K","model":"RefineNet","rank_in_archive_order":217,"of":235,"metrics":{"Validation mIoU":"40.7"},"uses_additional_data":false},{"leaderboard":"/sota/semantic-segmentation-on-ade20k-val","task":"Semantic Segmentation","dataset":"ADE20K val","model":"RefineNet (ResNet-152)","rank_in_archive_order":94,"of":95,"metrics":{"mIoU":"40.70"},"uses_additional_data":false},{"leaderboard":"/sota/semantic-segmentation-on-ade20k-val","task":"Semantic Segmentation","dataset":"ADE20K val","model":"RefineNet (ResNet-101)","rank_in_archive_order":95,"of":95,"metrics":{"mIoU":"40.20"},"uses_additional_data":false},{"leaderboard":"/sota/semantic-segmentation-on-coco-stuff-test","task":"Semantic Segmentation","dataset":"COCO-Stuff test","model":"RefineNet (ResNet-101)","rank_in_archive_order":19,"of":21,"metrics":{"mIoU":"33.6%"},"uses_additional_data":false},{"leaderboard":"/sota/semantic-segmentation-on-cityscapes","task":"Semantic Segmentation","dataset":"Cityscapes test","model":"RefineNet (ResNet-101)","rank_in_archive_order":67,"of":105,"metrics":{"Mean IoU (class)":"73.6%"},"uses_additional_data":false},{"leaderboard":"/sota/semantic-segmentation-on-nyu-depth-v2","task":"Semantic Segmentation","dataset":"NYU Depth v2","model":"RefineNet (ResNet-101)","rank_in_archive_order":89,"of":121,"metrics":{"Mean IoU":"46.5%"},"uses_additional_data":false},{"leaderboard":"/sota/semantic-segmentation-on-pascal-context","task":"Semantic Segmentation","dataset":"PASCAL Context","model":"RefineNet","rank_in_archive_order":54,"of":66,"metrics":{"mIoU":"47.3"},"uses_additional_data":false},{"leaderboard":"/sota/semantic-segmentation-on-pascal-voc-2012","task":"Semantic Segmentation","dataset":"PASCAL VOC 2012 test","model":"Multipath-RefineNet","rank_in_archive_order":14,"of":51,"metrics":{"Mean IoU":"84.2%"},"uses_additional_data":false},{"leaderboard":"/sota/semantic-segmentation-on-trans10k","task":"Semantic Segmentation","dataset":"Trans10K","model":"RefineNet","rank_in_archive_order":13,"of":15,"metrics":{"GFLOPs":"44.56","mIoU":"58.18%"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1611.06612","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1611.06612"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/hiyaroy12/multitask_learning","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/mindspore-ai/models/tree/master/research/cv/RefineNet","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/kingcong/RefineNet","reach":{"status":"gone","observed_at":"2026-09-18","how":"tree_404+repo_404"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/guosheng/refinenet","reach":{"status":"ok","spdx":"NOASSERTION"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/alililia/ascend_RefineNet","reach":{"status":"ok","spdx":"Apache-2.0"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/kingcong/gpu_RefineNet","reach":{"status":"gone","observed_at":"2026-09-18","how":"tree_404+repo_404"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/code-implementation1/Code7/tree/main/RefineNet","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/2023-MindSpore-1/ms-code-216/tree/main/RefineNet","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/2023-MindSpore-4/Code6/tree/main/RefineNet","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/MindSpore-paper-code-3/code5/tree/main/RefineNet","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/kdha0727/cancer-instance-segmentation-from-tissue","reach":{"status":"unanswered"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/yangyucheng000/RefineNet","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/hugozanini/realtime-semantic-segmentation","reach":{"status":"ok","spdx":"Apache-2.0"}}],"summary":{"ran_honours":1,"unverified":6},"by_repo_kind":{"listed":{"samples":7,"ran":1,"repositories":2}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":1,"samples":[{"code_sha256_prefix":"26490223efed770e","entry":"process_seg","repo":"hiyaroy12/multitask_learning","repo_kind":"listed","path":"refinenet-image-segmentation/RefineNet/test_captioning.py","file_url":"https://github.com/hiyaroy12/multitask_learning/blob/HEAD/refinenet-image-segmentation/RefineNet/test_captioning.py","link_basis":"first_harvest_node","language":"python","status":"ran_honours","verification_level":1,"contract_check":"HONOURS","metamorphic_tier":"invariant","behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"26490223efed770e"}},{"code_sha256_prefix":"5e57042613d98374","entry":"cal_hist","repo":"alililia/ascend_RefineNet","repo_kind":"listed","path":"postprocess.py","file_url":"https://github.com/alililia/ascend_RefineNet/blob/HEAD/postprocess.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"5e57042613d98374"}},{"code_sha256_prefix":"b13acc0922dade5c","entry":"get_img_size","repo":"alililia/ascend_RefineNet","repo_kind":"listed","path":"postprocess.py","file_url":"https://github.com/alililia/ascend_RefineNet/blob/HEAD/postprocess.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"b13acc0922dade5c"}},{"code_sha256_prefix":"ca5006a85f504b0d","entry":"get_lr","repo":"alililia/ascend_RefineNet","repo_kind":"listed","path":"src/learning_rates.py","file_url":"https://github.com/alililia/ascend_RefineNet/blob/HEAD/src/learning_rates.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"ca5006a85f504b0d"}},{"code_sha256_prefix":"874e384c8b2bd6bf","entry":"get_resized_size","repo":"alililia/ascend_RefineNet","repo_kind":"listed","path":"postprocess.py","file_url":"https://github.com/alililia/ascend_RefineNet/blob/HEAD/postprocess.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"874e384c8b2bd6bf"}},{"code_sha256_prefix":"9da7260d717664f9","entry":"linear_warmup_lr","repo":"alililia/ascend_RefineNet","repo_kind":"listed","path":"src/learning_rates.py","file_url":"https://github.com/alililia/ascend_RefineNet/blob/HEAD/src/learning_rates.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"9da7260d717664f9"}},{"code_sha256_prefix":"d4a0bdf1190319c4","entry":"warmup_cosine_annealing_lr","repo":"alililia/ascend_RefineNet","repo_kind":"listed","path":"src/learning_rates.py","file_url":"https://github.com/alililia/ascend_RefineNet/blob/HEAD/src/learning_rates.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"d4a0bdf1190319c4"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}