{"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/revisiting-rcnn-on-awakening-the","title":"Revisiting RCNN: On Awakening the Classification Power of Faster RCNN","arxiv_id":"1803.06799","date":"2018-03-19","proceeding":"ECCV 2018 9","authors":["Bowen Cheng","Yunchao Wei","Honghui Shi","Rogerio Feris","JinJun Xiong","Thomas Huang"],"abstract":"Recent region-based object detectors are usually built with separate\nclassification and localization branches on top of shared feature extraction\nnetworks. In this paper, we analyze failure cases of state-of-the-art detectors\nand observe that most hard false positives result from classification instead\nof localization. We conjecture that: (1) Shared feature representation is not\noptimal due to the mismatched goals of feature learning for classification and\nlocalization; (2) multi-task learning helps, yet optimization of the multi-task\nloss may result in sub-optimal for individual tasks; (3) large receptive field\nfor different scales leads to redundant context information for small\nobjects.We demonstrate the potential of detector classification power by a\nsimple, effective, and widely-applicable Decoupled Classification Refinement\n(DCR) network. DCR samples hard false positives from the base classifier in\nFaster RCNN and trains a RCNN-styled strong classifier. Experiments show new\nstate-of-the-art results on PASCAL VOC and COCO without any bells and whistles.","url_abs":"http://arxiv.org/abs/1803.06799v3","url_pdf":"http://arxiv.org/pdf/1803.06799v3.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":"revisiting-rcnn-on-awakening-the","repo_url":"https://github.com/SHI-Labs/Decoupled-Classification-Refinement","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"mxnet","reach":null},{"paper_slug":"revisiting-rcnn-on-awakening-the","repo_url":"https://github.com/bowenc0221/Decoupled-Classification-Refinement","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"revisiting-rcnn-on-awakening-the","repo_url":"https://github.com/makefile/DCR","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"revisiting-rcnn-on-awakening-the","repo_url":"https://github.com/MindSpore-paper-code-3/code10/tree/main/faster_rcnn_ssod","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"mindspore","reach":null},{"paper_slug":"revisiting-rcnn-on-awakening-the","repo_url":"https://github.com/MindSpore-paper-code-3/code3/tree/main/faster_rcnn_ssod","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"mindspore","reach":null},{"paper_slug":"revisiting-rcnn-on-awakening-the","repo_url":"https://github.com/MindSpore-paper-code-3/code7/tree/main/faster_rcnn_ssod","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"mindspore","reach":null},{"paper_slug":"revisiting-rcnn-on-awakening-the","repo_url":"https://github.com/MindSpore-paper-code-3/code8/tree/main/faster_rcnn_ssod","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"mindspore","reach":null}],"tasks":[{"task_slug":"classification-1","task_name":"Classification"},{"task_slug":"classification","task_name":"General Classification"},{"task_slug":"multi-task-learning","task_name":"Multi-Task Learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1803.06799","atlas_url":"https://app.syntology.ai/?focus=1803.06799","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1803.06799"}},"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/MindSpore-paper-code-3/code10/tree/main/faster_rcnn_ssod","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/MindSpore-paper-code-3/code3/tree/main/faster_rcnn_ssod","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/MindSpore-paper-code-3/code8/tree/main/faster_rcnn_ssod","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/MindSpore-paper-code-3/code7/tree/main/faster_rcnn_ssod","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/makefile/DCR","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/SHI-Labs/Decoupled-Classification-Refinement","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/bowenc0221/Decoupled-Classification-Refinement","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran_draft_wrong":3,"unverified":2},"by_repo_kind":{"listed":{"samples":5,"ran":3,"repositories":3}},"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":"015b5733d3e0b8a6","entry":"sample_rcnn","repo":"SHI-Labs/Decoupled-Classification-Refinement","repo_kind":"listed","path":"faster_rcnn_dcr/core/dcr.py","file_url":"https://github.com/SHI-Labs/Decoupled-Classification-Refinement/blob/HEAD/faster_rcnn_dcr/core/dcr.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"015b5733d3e0b8a6"}},{"code_sha256_prefix":"20ad0fa4e4fec335","entry":"sample_rois_hard_fg","repo":"SHI-Labs/Decoupled-Classification-Refinement","repo_kind":"listed","path":"faster_rcnn_dcr/core/dcr.py","file_url":"https://github.com/SHI-Labs/Decoupled-Classification-Refinement/blob/HEAD/faster_rcnn_dcr/core/dcr.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"20ad0fa4e4fec335"}},{"code_sha256_prefix":"6a2e623b98652915","entry":"sample_rois_random","repo":"SHI-Labs/Decoupled-Classification-Refinement","repo_kind":"listed","path":"faster_rcnn_dcr/core/dcr.py","file_url":"https://github.com/SHI-Labs/Decoupled-Classification-Refinement/blob/HEAD/faster_rcnn_dcr/core/dcr.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"6a2e623b98652915"}},{"code_sha256_prefix":"bbf8005ce03c08cf","entry":"get_rcnn_names","repo":"bowenc0221/Decoupled-Classification-Refinement","repo_kind":"listed","path":"faster_rcnn/core/metric.py","file_url":"https://github.com/bowenc0221/Decoupled-Classification-Refinement/blob/HEAD/faster_rcnn/core/metric.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":"bbf8005ce03c08cf"}},{"code_sha256_prefix":"d556ecc93005b402","entry":"get_rcnn_names","repo":"makefile/DCR","repo_kind":"listed","path":"faster_rcnn_dcr/core/metric.py","file_url":"https://github.com/makefile/DCR/blob/HEAD/faster_rcnn_dcr/core/metric.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":"d556ecc93005b402"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}