{"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/enhanced-training-of-query-based-object","title":"Enhanced Training of Query-Based Object Detection via Selective Query Recollection","arxiv_id":"2212.07593","date":"2022-12-15","proceeding":"CVPR 2023 1","authors":["Fangyi Chen","Han Zhang","Kai Hu","Yu-Kai Huang","Chenchen Zhu","Marios Savvides"],"abstract":"This paper investigates a phenomenon where query-based object detectors mispredict at the last decoding stage while predicting correctly at an intermediate stage. We review the training process and attribute the overlooked phenomenon to two limitations: lack of training emphasis and cascading errors from decoding sequence. We design and present Selective Query Recollection (SQR), a simple and effective training strategy for query-based object detectors. It cumulatively collects intermediate queries as decoding stages go deeper and selectively forwards the queries to the downstream stages aside from the sequential structure. Such-wise, SQR places training emphasis on later stages and allows later stages to work with intermediate queries from earlier stages directly. SQR can be easily plugged into various query-based object detectors and significantly enhances their performance while leaving the inference pipeline unchanged. As a result, we apply SQR on Adamixer, DAB-DETR, and Deformable-DETR across various settings (backbone, number of queries, schedule) and consistently brings 1.4-2.8 AP improvement.","url_abs":"https://arxiv.org/abs/2212.07593v3","url_pdf":"https://arxiv.org/pdf/2212.07593v3.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":"enhanced-training-of-query-based-object","repo_url":"https://github.com/fangyi-chen/sqr","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"enhanced-training-of-query-based-object","repo_url":"https://github.com/IDEA-Research/detrex","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"attribute","task_name":"Attribute"},{"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":"SQR-Adamixer-R101","rank_in_archive_order":17,"of":33,"metrics":{"AP":"49.8"},"uses_additional_data":false},{"leaderboard":"/sota/object-detection-on-coco-2017-val","task":"Object Detection","dataset":"COCO 2017 val","model":"SQR-Adamixer-R50","rank_in_archive_order":20,"of":33,"metrics":{"AP":"48.9"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/2212.07593","atlas_url":"https://app.syntology.ai/?focus=2212.07593","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}