ETRI-Knowledge Sharing Plaform

ENGLISH

성과물

논문 검색
구분 SCI
연도 ~ 키워드

상세정보

학술대회 Learning to Discriminate Information for Online Action Detection
Cited 11 time in scopus Download 11 time Share share facebook twitter linkedin kakaostory
저자
은현준, 문진영, 박종열, 정찬호, 김창익
발행일
202006
출처
Conference on Computer Vision and Pattern Recognition (CVPR) 2020, pp.806-815
DOI
https://dx.doi.org/10.1109/CVPR42600.2020.00089
협약과제
19HS3400, (딥뷰-1세부) 실시간 대규모 영상 데이터 이해·예측을 위한 고성능 비주얼 디스커버리 플랫폼 개발, 박종열
초록
From a streaming video, online action detection aims to identify actions in the present. For this task, previous methods use recurrent networks to model the temporal sequence of current action frames. However, these methods overlook the fact that an input image sequence includes background and irrelevant actions as well as the action of interest. For online action detection, in this paper, we propose a novel recurrent unit to explicitly discriminate the information relevant to an ongoing action from others. Our unit, named Information Discrimination Unit (IDU), decides whether to accumulate input information based on its relevance to the current action. This enables our recurrent network with IDU to learn a more discriminative representation for identifying ongoing actions. In experiments on two benchmark datasets, TVSeries and THUMOS-14, the proposed method outperforms state-of-the-art methods by a significant margin. Moreover, we demonstrate the effectiveness of our recurrent unit by conducting comprehensive ablation studies.
KSP 제안 키워드
Benchmark datasets, Discriminative representation, Image sequence, Input information, Online Action Detection, Recurrent network, Streaming video, Temporal sequence, state-of-The-Art