Papers › R²-Tuning: Efficient Image-to-Video Transfer Learning for Video Temporal Grounding

R²-Tuning: Efficient Image-to-Video Transfer Learning for Video Temporal Grounding

31 Mar 2024arXiv:2404.00801archive 2025-07-28

Ye Liu, Jixuan He, Wanhua Li, Junsik Kim, Donglai Wei, Hanspeter Pfister, Chang Wen Chen

Video temporal grounding (VTG) is a fine-grained video understanding problem that aims to ground relevant clips in untrimmed videos given natural language queries. Most existing VTG models are built upon frame-wise final-layer CLIP features, aided by additional temporal backbones (e.g., SlowFast) with sophisticated temporal reasoning mechanisms. In this work, we claim that CLIP itself already shows great potential for fine-grained spatial-temporal modeling, as each layer offers distinct yet useful information under different granularity levels. Motivated by this, we propose Reversed Recurrent Tuning (R²-Tuning), a parameter- and memory-efficient transfer learning framework for video temporal grounding. Our method learns a lightweight R² Block containing only 1.5% of the total parameters to perform progressive spatial-temporal modeling. Starting from the last layer of CLIP, R² Block recurrently aggregates spatial features from earlier layers, then refines temporal correlation conditioning on the given query, resulting in a coarse-to-fine scheme. R²-Tuning achieves state-of-the-art performance across three VTG tasks (i.e., moment retrieval, highlight detection, and video summarization) on six public benchmarks (i.e., QVHighlights, Charades-STA, Ego4D-NLQ, TACoS, YouTube Highlights, and TVSum) even without the additional backbone, demonstrating the significance and effectiveness of the proposed scheme. Our code is available at https://github.com/yeliudev/R2-Tuning.

PaperPDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

yeliudev/R2-Tuning officialmentioned in papermentioned on GitHubpytorchBSD-3-Clause report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Highlight DetectionMoment RetrievalNatural Language QueriesTransfer LearningVideo SummarizationVideo Understanding

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Highlight Detection QVHighlights R^2-Tuning Hit@1 64.20 #6 of 21 Archive leaderboard report
Highlight Detection QVHighlights R^2-Tuning mAP 40.75 #6 of 21 Archive leaderboard report
Moment Retrieval QVHighlights R^2-Tuning R@1 IoU=0.5 68.03 #12 of 32 Archive leaderboard report
Moment Retrieval QVHighlights R^2-Tuning R@1 IoU=0.7 49.35 #12 of 32 Archive leaderboard report
Moment Retrieval QVHighlights R^2-Tuning mAP 46.17 #12 of 32 Archive leaderboard report
Moment Retrieval QVHighlights R^2-Tuning mAP@0.5 69.04 #12 of 32 Archive leaderboard report
Moment Retrieval QVHighlights R^2-Tuning mAP@0.75 47.56 #12 of 32 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

CLIP

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections