Multi-shot temporal event localization: a benchmark

Conference item

Current developments in temporal event or action localization usually target actions captured by a single camera. However, extensive events or actions in the wild may be captured as a sequence of shots by multiple cameras at different positions. In this paper, we propose a new and challenging task called multi-shot temporal event localization, and accordingly, collect a large-scale dataset called MUlti-Shot EventS (MUSES). MUSES has 31,477 event instances for a total of 716 video hours. The core nature of MUSES is the frequent shot cuts, for an average of 19 shots per instance and 176 shots per video, which induces large intra-instance variations. Our comprehensive evaluations show that the state-of-the-art method in temporal action localization only achieves an mAP of 13.1% at IoU=0.5. As a minor contribution, we present a simple baseline approach for handling the intra-instance variations, which reports an mAP of 18.9% on MUSES and 56.9% on THUMOS14 at IoU=0.5. To facilitate research in this direction, we release the dataset and the project code at https://songbai.site/muses/.

Authors: Liu, X; Hu, Y; Bai, S; Ding, F; Bai, X

• Publication status: Published
• Peer review status: Peer reviewed
• Publisher: IEEE
• Host title: 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
• Pages: 12591-12601
• Publication date: 2021-11-13
• Acceptance date: 2021-03-01
• Event title: Conference on Computer Vision and Pattern Recognition (CVPR 2021)
• Event location: Virtual event
• Event website: http://cvpr2021.thecvf.com/
• Event start date: 2021-06-19
• Event end date: 2021-06-25
• DOI: 10.1109/CVPR46437.2021.01241
• EISSN: 2575-7075
• ISSN: 1063-6919
• EISBN: 9781665445092
• ISBN: 9781665445108
• Language: English
• Keywords: FFR