AI Paper Atlas

Research map / Video / Video-language understanding

The Devil is in the Spurious Correlations: Boosting Moment Retrieval with Dynamic Learning

Xinyang Zhou, Fanyue Wei, Lixin Duan, Angela Yao, Wen Li

ICCV 2025 · Video › Video-language understanding · approach: Video moment retrieval · 5 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add video synthesizer for dynamic context (VSDC) (ablation)QD-DETR · QVHighlights valmAP average: 42.54 -> 46.39Improved
add temporal dynamics enhancement module (TDEM) (ablation)QD-DETR · QVHighlights valmAP average: 42.54 -> 44.84Improved
combine VSDC and TDEM (ablation)QD-DETR · QVHighlights valR1@0.7: 46.77 -> 53.67Improved
combine VSDC and TDEM (ablation)QD-DETR · QVHighlights valSpurious R1@0.7: 9.35 -> 8.26Improved
select similar video pairs instead of random pairs for synthesis (ablation)QD-DETR · QVHighlights valmAP average: 47.56 -> 49.05Improved
add TD-DETR to BAM-DETRBAM-DETR · QVHighlights val (spurious)Spurious R1@0.7: 7.16 -> 1.61Improved
add TD-DETR to CG-DETRCG-DETR · QVHighlights val (spurious)Spurious R1@0.7: 4.65 -> 2.58Improved
add TD-DETR to QD-DETRQD-DETR · QVHighlights valmAP@0.75: 41.82 -> 49.86Improved
add TD-DETR to BAM-DETRBAM-DETR · QVHighlights valmAP average: 47.61 -> 48.82Improved
add TD-DETR to SnAGSnAG · Charades-STA testR1@0.7: 37.32 -> 42.35Improved
set sampling ratio alpha to 0.7 rather than 0.0 (ablation)TD-DETR · QVHighlights valmAP average: 10.09 -> 49.05Improved
set beta to 0.7 rather than 0.9 (less dynamics) (ablation)TD-DETR · QVHighlights valmAP average: 47.83 -> 49.05Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy