{"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/learning-pixel-level-semantic-affinity-with","title":"Learning Pixel-level Semantic Affinity with Image-level Supervision for Weakly Supervised Semantic Segmentation","arxiv_id":"1803.10464","date":"2018-03-28","proceeding":"CVPR 2018 6","authors":["Jiwoon Ahn","Suha Kwak"],"abstract":"The deficiency of segmentation labels is one of the main obstacles to\nsemantic segmentation in the wild. To alleviate this issue, we present a novel\nframework that generates segmentation labels of images given their image-level\nclass labels. In this weakly supervised setting, trained models have been known\nto segment local discriminative parts rather than the entire object area. Our\nsolution is to propagate such local responses to nearby areas which belong to\nthe same semantic entity. To this end, we propose a Deep Neural Network (DNN)\ncalled AffinityNet that predicts semantic affinity between a pair of adjacent\nimage coordinates. The semantic propagation is then realized by random walk\nwith the affinities predicted by AffinityNet. More importantly, the supervision\nemployed to train AffinityNet is given by the initial discriminative part\nsegmentation, which is incomplete as a segmentation annotation but sufficient\nfor learning semantic affinities within small image areas. Thus the entire\nframework relies only on image-level class labels and does not require any\nextra data or annotations. On the PASCAL VOC 2012 dataset, a DNN learned with\nsegmentation labels generated by our method outperforms previous models trained\nwith the same level of supervision, and is even as competitive as those relying\non stronger supervision.","url_abs":"http://arxiv.org/abs/1803.10464v2","url_pdf":"http://arxiv.org/pdf/1803.10464v2.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":"learning-pixel-level-semantic-affinity-with","repo_url":"https://github.com/jiwoon-ahn/psa","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"learning-pixel-level-semantic-affinity-with","repo_url":"https://github.com/suyukun666/CDA","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"segmentation","task_name":"Segmentation"},{"task_slug":"semantic-segmentation","task_name":"Semantic Segmentation"},{"task_slug":"weakly-supervised-semantic-segmentation-1","task_name":"Weakly supervised Semantic Segmentation"},{"task_slug":"weakly-supervised-semantic-segmentation","task_name":"Weakly-Supervised Semantic Segmentation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1803.10464","atlas_url":"https://app.syntology.ai/?focus=1803.10464","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1803.10464"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-25T09:33:49+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/suyukun666/CDA","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/jiwoon-ahn/psa","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"unverified":7},"by_repo_kind":{"official":{"samples":7,"ran":0,"repositories":1}},"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":0,"samples":[{"code_sha256_prefix":"fb4d78fd5d706f4e","entry":"convert_caffe_to_torch","repo":"jiwoon-ahn/psa","repo_kind":"official","path":"network/vgg16d.py","file_url":"https://github.com/jiwoon-ahn/psa/blob/HEAD/network/vgg16d.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"fb4d78fd5d706f4e"}},{"code_sha256_prefix":"95dfca01d7656b7c","entry":"convert_mxnet_to_torch","repo":"jiwoon-ahn/psa","repo_kind":"official","path":"network/resnet38d.py","file_url":"https://github.com/jiwoon-ahn/psa/blob/HEAD/network/resnet38d.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"95dfca01d7656b7c"}},{"code_sha256_prefix":"1da77dcda491ece7","entry":"crop_with_box","repo":"jiwoon-ahn/psa","repo_kind":"official","path":"tool/imutils.py","file_url":"https://github.com/jiwoon-ahn/psa/blob/HEAD/tool/imutils.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"1da77dcda491ece7"}},{"code_sha256_prefix":"9da0a975b889ff9b","entry":"get_indices_in_radius","repo":"jiwoon-ahn/psa","repo_kind":"official","path":"infer_aff.py","file_url":"https://github.com/jiwoon-ahn/psa/blob/HEAD/infer_aff.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"9da0a975b889ff9b"}},{"code_sha256_prefix":"3f83348d3b3bc43a","entry":"get_indices_of_pairs","repo":"jiwoon-ahn/psa","repo_kind":"official","path":"tool/pyutils.py","file_url":"https://github.com/jiwoon-ahn/psa/blob/HEAD/tool/pyutils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"3f83348d3b3bc43a"}},{"code_sha256_prefix":"14cb2e4c07b535f6","entry":"get_random_crop_box","repo":"jiwoon-ahn/psa","repo_kind":"official","path":"tool/imutils.py","file_url":"https://github.com/jiwoon-ahn/psa/blob/HEAD/tool/imutils.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"14cb2e4c07b535f6"}},{"code_sha256_prefix":"fab5abda02bc5312","entry":"random_crop","repo":"jiwoon-ahn/psa","repo_kind":"official","path":"tool/imutils.py","file_url":"https://github.com/jiwoon-ahn/psa/blob/HEAD/tool/imutils.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"fab5abda02bc5312"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}