{"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/dynamic-head-unifying-object-detection-heads","title":"Dynamic Head: Unifying Object Detection Heads with Attentions","arxiv_id":"2106.08322","date":"2021-06-15","proceeding":"CVPR 2021 1","authors":["Xiyang Dai","Yinpeng Chen","Bin Xiao","Dongdong Chen","Mengchen Liu","Lu Yuan","Lei Zhang"],"abstract":"The complex nature of combining localization and classification in object detection has resulted in the flourished development of methods. Previous works tried to improve the performance in various object detection heads but failed to present a unified view. In this paper, we present a novel dynamic head framework to unify object detection heads with attentions. By coherently combining multiple self-attention mechanisms between feature levels for scale-awareness, among spatial locations for spatial-awareness, and within output channels for task-awareness, the proposed approach significantly improves the representation ability of object detection heads without any computational overhead. Further experiments demonstrate that the effectiveness and efficiency of the proposed dynamic head on the COCO benchmark. With a standard ResNeXt-101-DCN backbone, we largely improve the performance over popular object detectors and achieve a new state-of-the-art at 54.0 AP. Furthermore, with latest transformer backbone and extra data, we can push current best COCO result to a new record at 60.6 AP. The code will be released at https://github.com/microsoft/DynamicHead.","url_abs":"https://arxiv.org/abs/2106.08322v1","url_pdf":"https://arxiv.org/pdf/2106.08322v1.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":"dynamic-head-unifying-object-detection-heads","repo_url":"https://github.com/microsoft/DynamicHead","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"dynamic-head-unifying-object-detection-heads","repo_url":"https://github.com/Coldestadam/DynamicHead","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"dynamic-head-unifying-object-detection-heads","repo_url":"https://github.com/open-mmlab/mmdetection","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}}],"tasks":[{"task_slug":"object","task_name":"Object"},{"task_slug":"object-detection","task_name":"Object Detection"},{"task_slug":"object-detection-1","task_name":"object-detection"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/object-detection-on-coco-2017-val","task":"Object Detection","dataset":"COCO 2017 val","model":"DyHead (Swin-T, multi scale)","rank_in_archive_order":33,"of":33,"metrics":{"AP50":"68","AP75":"54.3","APL":"64.2"},"uses_additional_data":false},{"leaderboard":"/sota/object-detection-on-coco-minival","task":"Object Detection","dataset":"COCO minival","model":"DyHead (Swin-L, multi scale, self-training)","rank_in_archive_order":25,"of":220,"metrics":{"AP50":"78.2","APL":"74.2","box AP":"60.3"},"uses_additional_data":true},{"leaderboard":"/sota/object-detection-on-coco-minival","task":"Object Detection","dataset":"COCO minival","model":"DyHead (Swin-L, multi scale)","rank_in_archive_order":36,"of":220,"metrics":{"AP50":"76.8","APL":"73.2","APM":"62.2","APS":"44.5","box AP":"58.4"},"uses_additional_data":false},{"leaderboard":"/sota/object-detection-on-coco-minival","task":"Object Detection","dataset":"COCO minival","model":"DyHead (ResNet-101)","rank_in_archive_order":103,"of":220,"metrics":{"box AP":"46.5"},"uses_additional_data":false},{"leaderboard":"/sota/object-detection-on-coco-minival","task":"Object Detection","dataset":"COCO minival","model":"DyHead (ResNeXt-64x4d-101-DCN, multi scale)","rank_in_archive_order":220,"of":220,"metrics":{"APL":"66.3"},"uses_additional_data":false},{"leaderboard":"/sota/object-detection-on-coco","task":"Object Detection","dataset":"COCO test-dev","model":"DyHead (Swin-L, multi scale, self-training)","rank_in_archive_order":26,"of":225,"metrics":{"AP50":"78.5","AP75":"66.6","APL":"74.2","APM":"64.0","box mAP":"60.6"},"uses_additional_data":false},{"leaderboard":"/sota/object-detection-on-coco","task":"Object Detection","dataset":"COCO test-dev","model":"DyHead (Swin-L, multi scale)","rank_in_archive_order":34,"of":225,"metrics":{"AP50":"77.1","AP75":"64.5","APL":"72.8","APM":"62.0","box mAP":"58.7"},"uses_additional_data":false},{"leaderboard":"/sota/object-detection-on-coco","task":"Object Detection","dataset":"COCO test-dev","model":"DyHead (ResNeXt-64x4d-101-DCN, multi scale)","rank_in_archive_order":58,"of":225,"metrics":{"AP50":"72.1","AP75":"59.3","box mAP":"54"},"uses_additional_data":false},{"leaderboard":"/sota/object-detection-on-coco","task":"Object Detection","dataset":"COCO test-dev","model":"DyHead (ResNeXt-64x4d-101)","rank_in_archive_order":116,"of":225,"metrics":{"AP50":"65.7","AP75":"51.9","box mAP":"47.7"},"uses_additional_data":false},{"leaderboard":"/sota/object-detection-on-coco","task":"Object Detection","dataset":"COCO test-dev","model":"DyHead (ResNet-50)","rank_in_archive_order":165,"of":225,"metrics":{"AP50":"60.7","AP75":"46.8","box mAP":"43"},"uses_additional_data":false},{"leaderboard":"/sota/object-detection-on-coco-o","task":"Object Detection","dataset":"COCO-O","model":"DyHead\n(Swin-L)","rank_in_archive_order":9,"of":45,"metrics":{"Average mAP":"35.3","Effective Robustness":"10.00"},"uses_additional_data":false},{"leaderboard":"/sota/object-detection-on-coco-o","task":"Object Detection","dataset":"COCO-O","model":"DyHead\n(ResNet-50)","rank_in_archive_order":31,"of":45,"metrics":{"Average mAP":"19.3","Effective Robustness":"0.16"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2106.08322","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2106.08322"}},"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/microsoft/DynamicHead","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/open-mmlab/mmdetection","reach":{"status":"ok","spdx":"Apache-2.0"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/Coldestadam/DynamicHead","reach":null}],"summary":{"ran":5,"ran_fixture":1,"ran_draft_wrong":1,"unverified":2},"by_repo_kind":{"official":{"samples":3,"ran":2,"repositories":1},"listed":{"samples":6,"ran":5,"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":"43a848f530c2b68c","entry":"DyHead_Block","repo":"Coldestadam/DynamicHead","repo_kind":"listed","path":"torch/DyHead.py","file_url":"https://github.com/Coldestadam/DynamicHead/blob/HEAD/torch/DyHead.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"43a848f530c2b68c"}},{"code_sha256_prefix":"f1f13347d3ea0492","entry":"DyReLUA","repo":"Coldestadam/DynamicHead","repo_kind":"listed","path":"torch/DyHead.py","file_url":"https://github.com/Coldestadam/DynamicHead/blob/HEAD/torch/DyHead.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"f1f13347d3ea0492"}},{"code_sha256_prefix":"ae71a995161d732d","entry":"Scale_Aware_Layer","repo":"Coldestadam/DynamicHead","repo_kind":"listed","path":"torch/DyHead.py","file_url":"https://github.com/Coldestadam/DynamicHead/blob/HEAD/torch/DyHead.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"ae71a995161d732d"}},{"code_sha256_prefix":"0d6fb3b73f74caa3","entry":"Spatial_Aware_Layer","repo":"Coldestadam/DynamicHead","repo_kind":"listed","path":"torch/DyHead.py","file_url":"https://github.com/Coldestadam/DynamicHead/blob/HEAD/torch/DyHead.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"0d6fb3b73f74caa3"}},{"code_sha256_prefix":"4c367c8f60b22fe0","entry":"Task_Aware_Layer","repo":"Coldestadam/DynamicHead","repo_kind":"listed","path":"torch/DyHead.py","file_url":"https://github.com/Coldestadam/DynamicHead/blob/HEAD/torch/DyHead.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"4c367c8f60b22fe0"}},{"code_sha256_prefix":"f9fd6241d935f07b","entry":"window_partition","repo":"microsoft/DynamicHead","repo_kind":"official","path":"extra/swint.py","file_url":"https://github.com/microsoft/DynamicHead/blob/HEAD/extra/swint.py","link_basis":"harvester_set","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":"invariant","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"f9fd6241d935f07b"}},{"code_sha256_prefix":"fb32094c6dbece71","entry":"window_reverse","repo":"microsoft/DynamicHead","repo_kind":"official","path":"extra/swint.py","file_url":"https://github.com/microsoft/DynamicHead/blob/HEAD/extra/swint.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":"invariant","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"fb32094c6dbece71"}},{"code_sha256_prefix":"0b96184174adb44f","entry":"DyHead","repo":"Coldestadam/DynamicHead","repo_kind":"listed","path":"torch/DyHead.py","file_url":"https://github.com/Coldestadam/DynamicHead/blob/HEAD/torch/DyHead.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":"0b96184174adb44f"}},{"code_sha256_prefix":"2028ba00ac3a78ae","entry":"permute_and_flatten","repo":"microsoft/DynamicHead","repo_kind":"official","path":"extra/atss.py","file_url":"https://github.com/microsoft/DynamicHead/blob/HEAD/extra/atss.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":"2028ba00ac3a78ae"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}