Deep feature enhancing and selecting network for weakly supervised temporal action localization

Jiaruo Yu; Yongxin Ge; Xiaolei QinZiqiang LiSheng HuangFeiyu Chen

首页> 外文期刊>Journal of visual communication & image representation >Deep feature enhancing and selecting network for weakly supervised temporal action localization

【24h】

Deep feature enhancing and selecting network for weakly supervised temporal action localization

机译：Deep feature enhancing and selecting network for weakly supervised temporal action localization

获取原文

获取原文并翻译 | 示例

掌桥外文数据库（机构版） >>

开具论文收录证明 >>

文献代查 >>

页面导航

摘要
著录项
相关主题

摘要

Weakly supervised temporal action localization is a challenging computer vision problem that uses onlyvideo-level labels and lacks the supervision of temporal annotations. In this task, the majority of existingmethods usually identify the most discriminative snippets and ignore other relevant snippets. To addressthis problem, we propose a deep feature enhancing and selecting network. It generates multiple masks forboth capturing more complete temporal interval of actions and keeping its high classification accuracy. Afterthat, we further propose a novel selection strategy to balance the influence of multiple masks and improvethe model performance. In the experiments, we evaluate the proposed method on the THUMOS’14 andActivityNet datasets, and the results show the effectiveness of our approach for weakly supervised temporalaction localization.

著录项

来源
《Journal of visual communication & image representation》 |2021年第10期|103276.1-103276.8|共8页
作者
Jiaruo Yu; Yongxin Ge; Xiaolei QinZiqiang LiSheng HuangFeiyu Chen;
展开▼
作者单位

School of Big Data and software Engineering, Chongqing University, Chongqing, China;

National Center for Applied Mathematics in Chongqing, Chongqing, China;

展开▼
收录信息
原文格式 PDF
正文语种英语
中图分类
关键词
Weakly supervised; Temporal action localization; Deep learning;

Deep feature enhancing and selecting network for weakly supervised temporal action localization

摘要

著录项

相关主题

期刊订阅