Abstract
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 outper-form the best feature-based methods. Code is available at https://github.com/sming256/AdaTAD.
Original language | English (US) |
---|---|
Title of host publication | Proceedings - 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2024 |
Publisher | IEEE Computer Society |
Pages | 18591-18601 |
Number of pages | 11 |
ISBN (Electronic) | 9798350353006 |
DOIs | |
State | Published - 2024 |
Event | 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2024 - Seattle, United States Duration: Jun 16 2024 → Jun 22 2024 |
Publication series
Name | Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition |
---|---|
ISSN (Print) | 1063-6919 |
Conference
Conference | 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2024 |
---|---|
Country/Territory | United States |
City | Seattle |
Period | 06/16/24 → 06/22/24 |
Bibliographical note
Publisher Copyright:© 2024 IEEE.
Keywords
- End-to-End Training
- Temporal Action Detection
- Video Understanding
ASJC Scopus subject areas
- Software
- Computer Vision and Pattern Recognition