{"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/deim-detr-with-improved-matching-for-fast","title":"DEIM: DETR with Improved Matching for Fast Convergence","arxiv_id":"2412.04234","date":"2024-12-05","proceeding":"CVPR 2025 1","authors":["Shihua Huang","Zhichao Lu","Xiaodong Cun","Yongjun Yu","Xiao Zhou","Xi Shen"],"abstract":"We introduce DEIM, an innovative and efficient training framework designed to accelerate convergence in real-time object detection with Transformer-based architectures (DETR). To mitigate the sparse supervision inherent in one-to-one (O2O) matching in DETR models, DEIM employs a Dense O2O matching strategy. This approach increases the number of positive samples per image by incorporating additional targets, using standard data augmentation techniques. While Dense O2O matching speeds up convergence, it also introduces numerous low-quality matches that could affect performance. To address this, we propose the Matchability-Aware Loss (MAL), a novel loss function that optimizes matches across various quality levels, enhancing the effectiveness of Dense O2O. Extensive experiments on the COCO dataset validate the efficacy of DEIM. When integrated with RT-DETR and D-FINE, it consistently boosts performance while reducing training time by 50%. Notably, paired with RT-DETRv2, DEIM achieves 53.2% AP in a single day of training on an NVIDIA 4090 GPU. Additionally, DEIM-trained real-time models outperform leading real-time object detectors, with DEIM-D-FINE-L and DEIM-D-FINE-X achieving 54.7% and 56.5% AP at 124 and 78 FPS on an NVIDIA T4 GPU, respectively, without the need for additional data. We believe DEIM sets a new baseline for advancements in real-time object detection. Our code and pre-trained models are available at https://github.com/ShihuaHuang95/DEIM.","url_abs":"https://arxiv.org/abs/2412.04234v3","url_pdf":"https://arxiv.org/pdf/2412.04234v3.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":"deim-detr-with-improved-matching-for-fast","repo_url":"https://github.com/shihuahuang95/deim","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"NOASSERTION"}}],"tasks":[{"task_slug":"data-augmentation","task_name":"Data Augmentation"},{"task_slug":null,"task_name":"GPU"},{"task_slug":"object-detection","task_name":"Object Detection"},{"task_slug":"real-time-object-detection","task_name":"Real-Time Object Detection"},{"task_slug":"object-detection-1","task_name":"object-detection"}],"methods":[{"method_slug":"absolute-position-encodings","method_name":"Absolute Position Encodings"},{"method_slug":"adam","method_name":"Adam"},{"method_slug":"attention","method_name":"Attention"},{"method_slug":"bpe","method_name":"BPE"},{"method_slug":"convolution","method_name":"Convolution"},{"method_slug":"dense-connections","method_name":"Dense Connections"},{"method_slug":"detr","method_name":"Detr"},{"method_slug":"dropout","method_name":"Dropout"},{"method_slug":"feedforward-network","method_name":"Feedforward Network"},{"method_slug":"label-smoothing","method_name":"Label Smoothing"},{"method_slug":"layer-normalization","method_name":"Layer Normalization"},{"method_slug":"linear-layer","method_name":"Linear Layer"},{"method_slug":"multi-head-attention","method_name":"Multi-Head Attention"},{"method_slug":"position-wise-feed-forward-layer","method_name":"Position-Wise Feed-Forward Layer"},{"method_slug":"residual-connection","method_name":"Residual Connection"},{"method_slug":"softmax","method_name":"Softmax"},{"method_slug":"transformer","method_name":"Transformer"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/real-time-object-detection-on-coco","task":"Real-Time Object Detection","dataset":"COCO (Common Objects in Context)","model":"DEIM-D-FINE-X+","rank_in_archive_order":1,"of":82,"metrics":{"FPS (V100, b=1)":"78 (T4)","box AP":"59.5"},"uses_additional_data":false},{"leaderboard":"/sota/real-time-object-detection-on-coco","task":"Real-Time Object Detection","dataset":"COCO (Common Objects in Context)","model":"DEIM-D-FINE-X","rank_in_archive_order":8,"of":82,"metrics":{"FPS (V100, b=1)":"78 (T4)","box AP":"56.5"},"uses_additional_data":false},{"leaderboard":"/sota/real-time-object-detection-on-coco","task":"Real-Time Object Detection","dataset":"COCO (Common Objects in Context)","model":"DEIM-D-FINE-L","rank_in_archive_order":20,"of":82,"metrics":{"FPS (V100, b=1)":"124 (T4)","box AP":"54.7"},"uses_additional_data":false},{"leaderboard":"/sota/real-time-object-detection-on-coco","task":"Real-Time Object Detection","dataset":"COCO (Common Objects in Context)","model":"DEIM-D-FINE-M","rank_in_archive_order":39,"of":82,"metrics":{"FPS (V100, b=1)":"178 (T4)","box AP":"52.7"},"uses_additional_data":false},{"leaderboard":"/sota/real-time-object-detection-on-coco","task":"Real-Time Object Detection","dataset":"COCO (Common Objects in Context)","model":"DEIM-D-FINE-S","rank_in_archive_order":60,"of":82,"metrics":{"FPS (V100, b=1)":"287 (T4)","box AP":"49.0"},"uses_additional_data":false},{"leaderboard":"/sota/real-time-object-detection-on-coco","task":"Real-Time Object Detection","dataset":"COCO (Common Objects in Context)","model":"DEIM-D-FINE-X+","rank_in_archive_order":82,"of":82,"metrics":{"FPS (V100, b=1)":"78 (T4)"},"uses_additional_data":true}],"syntology":{"syntology_url":"https://syntology.ai/paper/2412.04234","atlas_url":"https://app.syntology.ai/?focus=2412.04234","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2412.04234"}},"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/shihuahuang95/deim","reach":{"status":"ok","spdx":"NOASSERTION"}}],"summary":{"ran_honours":1,"unverified":12},"by_repo_kind":{"official":{"samples":13,"ran":1,"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":13,"samples":[{"code_sha256_prefix":"988a3c854b1b13d0","entry":"autopad","repo":"shihuahuang95/deim","repo_kind":"official","path":"engine/backbone/csp_darknet.py","file_url":"https://github.com/shihuahuang95/deim/blob/HEAD/engine/backbone/csp_darknet.py","link_basis":"harvester_set","language":"python","status":"ran_honours","verification_level":1,"contract_check":"HONOURS","metamorphic_tier":"well_formed","behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"988a3c854b1b13d0"}},{"code_sha256_prefix":"8c4a2df028bbf447","entry":"bias_init_with_prob","repo":"shihuahuang95/deim","repo_kind":"official","path":"engine/deim/utils.py","file_url":"https://github.com/shihuahuang95/deim/blob/HEAD/engine/deim/utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"8c4a2df028bbf447"}},{"code_sha256_prefix":"cb48d7481c17a1f7","entry":"box_cxcywh_to_xyxy","repo":"shihuahuang95/deim","repo_kind":"official","path":"engine/deim/box_ops.py","file_url":"https://github.com/shihuahuang95/deim/blob/HEAD/engine/deim/box_ops.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"cb48d7481c17a1f7"}},{"code_sha256_prefix":"253961917cbe7713","entry":"box_iou","repo":"shihuahuang95/deim","repo_kind":"official","path":"engine/deim/box_ops.py","file_url":"https://github.com/shihuahuang95/deim/blob/HEAD/engine/deim/box_ops.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"253961917cbe7713"}},{"code_sha256_prefix":"db44d81a87192d0f","entry":"box_xyxy_to_cxcywh","repo":"shihuahuang95/deim","repo_kind":"official","path":"engine/deim/box_ops.py","file_url":"https://github.com/shihuahuang95/deim/blob/HEAD/engine/deim/box_ops.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"db44d81a87192d0f"}},{"code_sha256_prefix":"9524bdc261cbc2dc","entry":"deformable_attention_core_func","repo":"shihuahuang95/deim","repo_kind":"official","path":"engine/deim/utils.py","file_url":"https://github.com/shihuahuang95/deim/blob/HEAD/engine/deim/utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"9524bdc261cbc2dc"}},{"code_sha256_prefix":"07ceddb7b30df7b3","entry":"distance2bbox","repo":"shihuahuang95/deim","repo_kind":"official","path":"engine/deim/dfine_utils.py","file_url":"https://github.com/shihuahuang95/deim/blob/HEAD/engine/deim/dfine_utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"07ceddb7b30df7b3"}},{"code_sha256_prefix":"4e994353f1d6e99a","entry":"get_contrastive_denoising_training_group","repo":"shihuahuang95/deim","repo_kind":"official","path":"engine/deim/denoising.py","file_url":"https://github.com/shihuahuang95/deim/blob/HEAD/engine/deim/denoising.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"4e994353f1d6e99a"}},{"code_sha256_prefix":"2fff86fc7c917b0c","entry":"inverse_sigmoid","repo":"shihuahuang95/deim","repo_kind":"official","path":"engine/deim/utils.py","file_url":"https://github.com/shihuahuang95/deim/blob/HEAD/engine/deim/utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"2fff86fc7c917b0c"}},{"code_sha256_prefix":"2a08c4c250501510","entry":"make_divisible","repo":"shihuahuang95/deim","repo_kind":"official","path":"engine/backbone/csp_darknet.py","file_url":"https://github.com/shihuahuang95/deim/blob/HEAD/engine/backbone/csp_darknet.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"2a08c4c250501510"}},{"code_sha256_prefix":"ec4284e2142c46f9","entry":"mod","repo":"shihuahuang95/deim","repo_kind":"official","path":"engine/deim/postprocessor.py","file_url":"https://github.com/shihuahuang95/deim/blob/HEAD/engine/deim/postprocessor.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"ec4284e2142c46f9"}},{"code_sha256_prefix":"0f787f3e8b97aba1","entry":"translate_gt","repo":"shihuahuang95/deim","repo_kind":"official","path":"engine/deim/dfine_utils.py","file_url":"https://github.com/shihuahuang95/deim/blob/HEAD/engine/deim/dfine_utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"0f787f3e8b97aba1"}},{"code_sha256_prefix":"1a45b1a466ee92a4","entry":"weighting_function","repo":"shihuahuang95/deim","repo_kind":"official","path":"engine/deim/dfine_utils.py","file_url":"https://github.com/shihuahuang95/deim/blob/HEAD/engine/deim/dfine_utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"1a45b1a466ee92a4"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}