Papers › Dual-path Adaptation from Image to Video Transformers

Dual-path Adaptation from Image to Video Transformers

17 Mar 2023CVPR 2023 1arXiv:2303.09857archive 2025-07-28

Jungin Park, Jiyoung Lee, Kwanghoon Sohn

In this paper, we efficiently transfer the surpassing representation power of the vision foundation models, such as ViT and Swin, for video understanding with only a few trainable parameters. Previous adaptation methods have simultaneously considered spatial and temporal modeling with a unified learnable module but still suffered from fully leveraging the representative capabilities of image transformers. We argue that the popular dual-path (two-stream) architecture in video models can mitigate this problem. We propose a novel DualPath adaptation separated into spatial and temporal adaptation paths, where a lightweight bottleneck adapter is employed in each transformer block. Especially for temporal dynamic modeling, we incorporate consecutive frames into a grid-like frameset to precisely imitate vision transformers' capability that extrapolates relationships between tokens. In addition, we extensively investigate the multiple baselines from a unified perspective in video understanding and compare them with DualPath. Experimental results on four action recognition benchmarks prove that pretrained image transformers with DualPath can be effectively generalized beyond the data domain.

PaperPDFConference PDFCode

In Syntology View this paper on Syntology: its repositories, every harvested function with whether it ran, its licence and the call to fetch it.

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

Code

park-jungin/dualpath officialmentioned in paperpytorch 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

Action ClassificationAction RecognitionAction Recognition In VideosActivity RecognitionActivity Recognition In VideosVideo Understanding

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Action Classification Diving-48 DualPath w/ ViT-B/16 Acc@1 88.7 #1 of 1 Archive leaderboard report
Action Classification HMDB51 DualPath w/ ViT-B/16 MLPs. Acc@1 75.6 #1 of 1 Archive leaderboard report
Action Classification Kinetics-400 DualPath w/ ViT-L/14 Acc@1 87.7 #33 of 207 Archive leaderboard report
Action Classification Kinetics-400 DualPath w/ ViT-L/14 Acc@5 97.8 #33 of 207 Archive leaderboard report
Action Classification Kinetics-400 DualPath w/ ViT-B/16 Acc@1 85.4 #53 of 207 Archive leaderboard report
Action Classification Kinetics-400 DualPath w/ ViT-B/16 Acc@5 97.1 #53 of 207 Archive leaderboard report
Action Recognition Diving-48 DUALPATH Accuracy 88.7 #4 of 18 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

Adapter

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