{"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/query-dependent-video-representation-for","title":"Query-Dependent Video Representation for Moment Retrieval and Highlight Detection","arxiv_id":"2303.13874","date":"2023-03-24","proceeding":"CVPR 2023 1","authors":["WonJun Moon","Sangeek Hyun","Sanguk Park","Dongchan Park","Jae-Pil Heo"],"abstract":"Recently, video moment retrieval and highlight detection (MR/HD) are being spotlighted as the demand for video understanding is drastically increased. The key objective of MR/HD is to localize the moment and estimate clip-wise accordance level, i.e., saliency score, to the given text query. Although the recent transformer-based models brought some advances, we found that these methods do not fully exploit the information of a given query. For example, the relevance between text query and video contents is sometimes neglected when predicting the moment and its saliency. To tackle this issue, we introduce Query-Dependent DETR (QD-DETR), a detection transformer tailored for MR/HD. As we observe the insignificant role of a given query in transformer architectures, our encoding module starts with cross-attention layers to explicitly inject the context of text query into video representation. Then, to enhance the model's capability of exploiting the query information, we manipulate the video-query pairs to produce irrelevant pairs. Such negative (irrelevant) video-query pairs are trained to yield low saliency scores, which in turn, encourages the model to estimate precise accordance between query-video pairs. Lastly, we present an input-adaptive saliency predictor which adaptively defines the criterion of saliency scores for the given video-query pairs. Our extensive studies verify the importance of building the query-dependent representation for MR/HD. Specifically, QD-DETR outperforms state-of-the-art methods on QVHighlights, TVSum, and Charades-STA datasets. Codes are available at github.com/wjun0830/QD-DETR.","url_abs":"https://arxiv.org/abs/2303.13874v1","url_pdf":"https://arxiv.org/pdf/2303.13874v1.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":"query-dependent-video-representation-for","repo_url":"https://github.com/wjun0830/qd-detr","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"highlight-detection","task_name":"Highlight Detection"},{"task_slug":"moment-retrieval","task_name":"Moment Retrieval"},{"task_slug":"natural-language-queries","task_name":"Natural Language Queries"},{"task_slug":"retrieval","task_name":"Retrieval"},{"task_slug":"video-grounding","task_name":"Video Grounding"},{"task_slug":"video-understanding","task_name":"Video Understanding"}],"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/highlight-detection-on-qvhighlights","task":"Highlight Detection","dataset":"QVHighlights","model":"QD-DETR","rank_in_archive_order":13,"of":21,"metrics":{"Hit@1":"62.87","mAP":"39.04"},"uses_additional_data":false},{"leaderboard":"/sota/highlight-detection-on-qvhighlights","task":"Highlight Detection","dataset":"QVHighlights","model":"QD-DETR (only Video)","rank_in_archive_order":14,"of":21,"metrics":{"Hit@1":"62.40","mAP":"38.94"},"uses_additional_data":false},{"leaderboard":"/sota/highlight-detection-on-qvhighlights","task":"Highlight Detection","dataset":"QVHighlights","model":"QD-DETR (w/ PT)","rank_in_archive_order":15,"of":21,"metrics":{"Hit@1":"62.27","mAP":"38.52"},"uses_additional_data":false},{"leaderboard":"/sota/highlight-detection-on-qvhighlights","task":"Highlight Detection","dataset":"QVHighlights","model":"QD-DETR (only Video w/ PT)","rank_in_archive_order":21,"of":21,"metrics":{"Hit@1":"61.91"},"uses_additional_data":false},{"leaderboard":"/sota/highlight-detection-on-tvsum","task":"Highlight Detection","dataset":"TvSum","model":"QD-DETR","rank_in_archive_order":4,"of":7,"metrics":{"mAP":"86.6"},"uses_additional_data":false},{"leaderboard":"/sota/highlight-detection-on-tvsum","task":"Highlight Detection","dataset":"TvSum","model":"QD-DETR (only Video)","rank_in_archive_order":6,"of":7,"metrics":{"mAP":"85.0"},"uses_additional_data":false},{"leaderboard":"/sota/moment-retrieval-on-charades-sta","task":"Moment Retrieval","dataset":"Charades-STA","model":"QD-DETR (Only Video)","rank_in_archive_order":18,"of":25,"metrics":{"R@1 IoU=0.5":"57.31","R@1 IoU=0.7":"32.55"},"uses_additional_data":false},{"leaderboard":"/sota/moment-retrieval-on-qvhighlights","task":"Moment Retrieval","dataset":"QVHighlights","model":"QD-DETR (w/ PT)","rank_in_archive_order":20,"of":32,"metrics":{"R@1 IoU=0.5":"64.1","R@1 IoU=0.7":"46.1","mAP":"40.62","mAP@0.5":"64.3","mAP@0.75":"40.5"},"uses_additional_data":false},{"leaderboard":"/sota/moment-retrieval-on-qvhighlights","task":"Moment Retrieval","dataset":"QVHighlights","model":"QD-DETR (w/ audio)","rank_in_archive_order":21,"of":32,"metrics":{"R@1 IoU=0.5":"63.06","R@1 IoU=0.7":"45.10","mAP":"40.19","mAP@0.5":"63.04","mAP@0.75":"40.10"},"uses_additional_data":false},{"leaderboard":"/sota/moment-retrieval-on-qvhighlights","task":"Moment Retrieval","dataset":"QVHighlights","model":"QD-DETR (only Video  w/ PT ASR Captions)","rank_in_archive_order":23,"of":32,"metrics":{"R@1 IoU=0.5":"63.2","R@1 IoU=0.7":"45.2","mAP":"40.0","mAP@0.5":"63.4","mAP@0.75":"40.4"},"uses_additional_data":false},{"leaderboard":"/sota/moment-retrieval-on-qvhighlights","task":"Moment Retrieval","dataset":"QVHighlights","model":"QD-DETR (only Video)","rank_in_archive_order":24,"of":32,"metrics":{"R@1 IoU=0.5":"62.40","R@1 IoU=0.7":"44.98","mAP":"39.86","mAP@0.5":"62.52","mAP@0.75":"39.88"},"uses_additional_data":false},{"leaderboard":"/sota/video-grounding-on-qvhighlights","task":"Video Grounding","dataset":"QVHighlights","model":"QD-DETR","rank_in_archive_order":4,"of":7,"metrics":{"R@1,IoU=0.5":"62.40","R@1,IoU=0.7":"44.98"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2303.13874","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2303.13874"}},"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/wjun0830/qd-detr","reach":null},{"provenance":"deterministic:regex_extraction","url":"https://github.com/wjun0830/QD-DETR","reach":null}],"summary":{"ran":1},"by_repo_kind":{"official":{"samples":1,"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":1,"samples":[{"code_sha256_prefix":"32a0a976c1ecbbd0","entry":"QDDETR","repo":"wjun0830/QD-DETR","repo_kind":"official","path":"qd_detr/model.py","file_url":"https://github.com/wjun0830/QD-DETR/blob/HEAD/qd_detr/model.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"32a0a976c1ecbbd0"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}