Papers › End-to-End Temporal Action Detection with 1B Parameters Across 1000 Frames

End-to-End Temporal Action Detection with 1B Parameters Across 1000 Frames

28 Nov 2023CVPR 2024 1arXiv:2311.17241archive 2025-07-28

Shuming Liu, Chen-Lin Zhang, Chen Zhao, Bernard Ghanem

Recently, temporal action detection (TAD) has seen significant performance improvement with end-to-end training. However, due to the memory bottleneck, only models with limited scales and limited data volumes can afford end-to-end training, which inevitably restricts TAD performance. In this paper, we reduce the memory consumption for end-to-end training, and manage to scale up the TAD backbone to 1 billion parameters and the input video to 1,536 frames, leading to significant detection performance. The key to our approach lies in our proposed temporal-informative adapter (TIA), which is a novel lightweight module that reduces training memory. Using TIA, we free the humongous backbone from learning to adapt to the TAD task by only updating the parameters in TIA. TIA also leads to better TAD representation by temporally aggregating context from adjacent frames throughout the backbone. We evaluate our model across four representative datasets. Owing to our efficient design, we are able to train end-to-end on VideoMAEv2-giant and achieve 75.4% mAP on THUMOS14, being the first end-to-end model to outperform the best feature-based methods. Code is available at https://github.com/sming256/AdaTAD.

PaperPDFConference PDFCode

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

Code

sming256/OpenTAD officialpytorch report
sming256/AdaTAD mentioned in papermentioned on GitHubpytorch 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 DetectionTemporal Action Localization

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Temporal Action Localization ActivityNet-1.3 AdaTAD (VideoMAEv2-giant) mAP 41.93 #4 of 33 Archive leaderboard report
Temporal Action Localization ActivityNet-1.3 AdaTAD (VideoMAEv2-giant) mAP IOU@0.5 61.72 #4 of 33 Archive leaderboard report
Temporal Action Localization ActivityNet-1.3 AdaTAD (VideoMAEv2-giant) mAP IOU@0.75 43.35 #4 of 33 Archive leaderboard report
Temporal Action Localization ActivityNet-1.3 AdaTAD (VideoMAEv2-giant) mAP IOU@0.95 10.85 #4 of 33 Archive leaderboard report
Temporal Action Localization EPIC-KITCHENS-100 AdaTAD (verb, VideoMAE-L) Avg mAP (0.1-0.5) 29.3 #1 of 6 Archive leaderboard report
Temporal Action Localization EPIC-KITCHENS-100 AdaTAD (verb, VideoMAE-L) mAP IOU@0.1 33.1 #1 of 6 Archive leaderboard report
Temporal Action Localization EPIC-KITCHENS-100 AdaTAD (verb, VideoMAE-L) mAP IOU@0.2 32.2 #1 of 6 Archive leaderboard report
Temporal Action Localization EPIC-KITCHENS-100 AdaTAD (verb, VideoMAE-L) mAP IOU@0.3 30.4 #1 of 6 Archive leaderboard report
Temporal Action Localization EPIC-KITCHENS-100 AdaTAD (verb, VideoMAE-L) mAP IOU@0.4 27.5 #1 of 6 Archive leaderboard report
Temporal Action Localization EPIC-KITCHENS-100 AdaTAD (verb, VideoMAE-L) mAP IOU@0.5 23.1 #1 of 6 Archive leaderboard report
Temporal Action Localization THUMOS’14 AdaTAD (VideoMAEv2-giant) Avg mAP (0.3:0.7) 76.9 #1 of 42 Archive leaderboard report
Temporal Action Localization THUMOS’14 AdaTAD (VideoMAEv2-giant) mAP IOU@0.3 89.7 #1 of 42 Archive leaderboard report
Temporal Action Localization THUMOS’14 AdaTAD (VideoMAEv2-giant) mAP IOU@0.4 86.7 #1 of 42 Archive leaderboard report
Temporal Action Localization THUMOS’14 AdaTAD (VideoMAEv2-giant) mAP IOU@0.5 80.9 #1 of 42 Archive leaderboard report
Temporal Action Localization THUMOS’14 AdaTAD (VideoMAEv2-giant) mAP IOU@0.6 71.0 #1 of 42 Archive leaderboard report
Temporal Action Localization THUMOS’14 AdaTAD (VideoMAEv2-giant) mAP IOU@0.7 56.1 #1 of 42 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