{"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-convolution-attention-over","title":"Dynamic Convolution: Attention over Convolution Kernels","arxiv_id":"1912.03458","date":"2019-12-07","proceeding":"CVPR 2020 6","authors":["Yinpeng Chen","Xiyang Dai","Mengchen Liu","Dong-Dong Chen","Lu Yuan","Zicheng Liu"],"abstract":"Light-weight convolutional neural networks (CNNs) suffer performance degradation as their low computational budgets constrain both the depth (number of convolution layers) and the width (number of channels) of CNNs, resulting in limited representation capability. To address this issue, we present Dynamic Convolution, a new design that increases model complexity without increasing the network depth or width. Instead of using a single convolution kernel per layer, dynamic convolution aggregates multiple parallel convolution kernels dynamically based upon their attentions, which are input dependent. Assembling multiple kernels is not only computationally efficient due to the small kernel size, but also has more representation power since these kernels are aggregated in a non-linear way via attention. By simply using dynamic convolution for the state-of-the-art architecture MobileNetV3-Small, the top-1 accuracy of ImageNet classification is boosted by 2.9% with only 4% additional FLOPs and 2.9 AP gain is achieved on COCO keypoint detection.","url_abs":"https://arxiv.org/abs/1912.03458v2","url_pdf":"https://arxiv.org/pdf/1912.03458v2.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-convolution-attention-over","repo_url":"https://github.com/TArdelean/DynamicConvolution","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"dynamic-convolution-attention-over","repo_url":"https://github.com/kaijieshi7/Dynamic-convolution-Pytorch","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}},{"paper_slug":"dynamic-convolution-attention-over","repo_url":"https://github.com/prstrive/CondConv-tensorflow","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"dynamic-convolution-attention-over","repo_url":"https://github.com/xmu-xiaoma666/External-Attention-pytorch","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"dynamic-convolution-attention-over","repo_url":"https://github.com/mindspore-courses/External-Attention-MindSpore/blob/main/model/conv/DynamicConv.py","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"mindspore","reach":null}],"tasks":[{"task_slug":"image-classification","task_name":"Image Classification"},{"task_slug":"keypoint-detection","task_name":"Keypoint Detection"}],"methods":[{"method_slug":"convolution","method_name":"Convolution"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/image-classification-on-imagenet","task":"Image Classification","dataset":"ImageNet","model":"DY-MobileNetV2 ×1.0","rank_in_archive_order":977,"of":1060,"metrics":{"GFLOPs":"0,626","Number of params":"11.1M","Top 1 Accuracy":"74.4%"},"uses_additional_data":false},{"leaderboard":"/sota/image-classification-on-imagenet","task":"Image Classification","dataset":"ImageNet","model":"DY-MobileNetV2 ×0.75","rank_in_archive_order":992,"of":1060,"metrics":{"GFLOPs":"0.435","Number of params":"7M","Top 1 Accuracy":"72.8%"},"uses_additional_data":false},{"leaderboard":"/sota/image-classification-on-imagenet","task":"Image Classification","dataset":"ImageNet","model":"DY-ResNet-18","rank_in_archive_order":993,"of":1060,"metrics":{"GFLOPs":"3.7","Number of params":"42.7M","Top 1 Accuracy":"72.7%"},"uses_additional_data":false},{"leaderboard":"/sota/image-classification-on-imagenet","task":"Image Classification","dataset":"ImageNet","model":"DY-MobileNetV3-Small","rank_in_archive_order":1025,"of":1060,"metrics":{"GFLOPs":"0.137","Number of params":"4.8M","Top 1 Accuracy":"69.7%"},"uses_additional_data":false},{"leaderboard":"/sota/image-classification-on-imagenet","task":"Image Classification","dataset":"ImageNet","model":"DY-MobileNetV2 ×0.5","rank_in_archive_order":1028,"of":1060,"metrics":{"GFLOPs":"0.203","Number of params":"4M","Top 1 Accuracy":"69.4%"},"uses_additional_data":false},{"leaderboard":"/sota/image-classification-on-imagenet","task":"Image Classification","dataset":"ImageNet","model":"DY-ResNet-10","rank_in_archive_order":1038,"of":1060,"metrics":{"GFLOPs":"1.82","Number of params":"18.6M","Top 1 Accuracy":"67.7%"},"uses_additional_data":false},{"leaderboard":"/sota/image-classification-on-imagenet","task":"Image Classification","dataset":"ImageNet","model":"DY-MobileNetV2 ×0.35","rank_in_archive_order":1048,"of":1060,"metrics":{"GFLOPs":"0.124","Number of params":"2.8M","Top 1 Accuracy":"64.9%"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1912.03458","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1912.03458"}},"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/mindspore-courses/External-Attention-MindSpore/blob/main/model/conv/DynamicConv.py","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/prstrive/CondConv-tensorflow","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/kaijieshi7/Dynamic-convolution-Pytorch","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/TArdelean/DynamicConvolution","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/xmu-xiaoma666/External-Attention-pytorch","reach":null}],"summary":{"ran_honours":1,"unverified":3},"by_repo_kind":{"listed":{"samples":4,"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":0,"samples":[{"code_sha256_prefix":"63ec083689f80f47","entry":"make_divisible","repo":"TArdelean/DynamicConvolution","repo_kind":"listed","path":"models/mobilenetv3.py","file_url":"https://github.com/TArdelean/DynamicConvolution/blob/HEAD/models/mobilenetv3.py","link_basis":"harvester_set","language":"python","status":"ran_honours","verification_level":1,"contract_check":"HONOURS","metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"63ec083689f80f47"}},{"code_sha256_prefix":"e36ca5f4167e4e43","entry":"compute_entropy","repo":"TArdelean/DynamicConvolution","repo_kind":"listed","path":"inspect_attention.py","file_url":"https://github.com/TArdelean/DynamicConvolution/blob/HEAD/inspect_attention.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":"e36ca5f4167e4e43"}},{"code_sha256_prefix":"2376fefdf0b2d1bc","entry":"get_cifar_dataset","repo":"prstrive/CondConv-tensorflow","repo_kind":"listed","path":"dataset.py","file_url":"https://github.com/prstrive/CondConv-tensorflow/blob/HEAD/dataset.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":"2376fefdf0b2d1bc"}},{"code_sha256_prefix":"f9cfc6fd89dbb714","entry":"get_datasets","repo":"prstrive/CondConv-tensorflow","repo_kind":"listed","path":"dataset.py","file_url":"https://github.com/prstrive/CondConv-tensorflow/blob/HEAD/dataset.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":"f9cfc6fd89dbb714"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}