arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2201.11345cs.CVcs.AI
收藏

探索全局多样性和局部上下文的视频摘要

Exploring Global Diversity and Local Context for Video Summarization

Hangzhou Dianzi University(杭州电子科技大学);University College London(伦敦大学学院);Fudan University(复旦大学);University of California, San Diego(加利福尼亚大学圣地亚哥分校)

Yingchao Pan, Ouhan Huang, Qinghao Ye, Zhongjin Li, Wenjiang Wang, Guodun Li, Yuxing Chen

摘要:

视频摘要的目的是自动生成一个多样的和简洁的摘要,这是有用的大规模视频处理。大多数方法倾向于采用跨视频帧的自注意力机制,未能对视频帧的多样性进行建模。为了解决这个问题,我们重新审视了自注意力机制中的成对相似性度量,发现现有的内积亲和力导致了区分性特征,而不是多样化特征。针对这一现象,我们提出了全局分散注意力算法,该算法使用平方欧几里德距离来计算亲和度。此外,我们的局部上下文注意力模型的局部上下文信息,以消除视频中的冗余。将这两种注意力机制结合起来,提出了一种基于多样化上下文注意机制的视频摘要化模型,即SUM-DCA。在基准数据集上进行了大量的实验,以验证SUM-DCA的有效性和优越性,在F分数和基于等级的评价,没有任何花里胡哨。

英文摘要:

Video summarization aims to automatically generate a diverse and concise summary which is useful in large-scale video processing. Most of the methods tend to adopt self-attention mechanism across video frames, which fails to model the diversity of video frames. To alleviate this problem, we revisit the pairwise similarity measurement in self-attention mechanism and find that the existing inner-product affinity leads to discriminative features rather than diversified features. In light of this phenomenon, we propose global diverse attention which uses the squared Euclidean distance instead to compute the affinities. Moreover, we model the local contextual information by novel local contextual attention to remove the redundancy in the video. By combining these two attention mechanisms, a video SUMmarization model with Diversified Contextual Attention scheme is developed, namely SUM-DCA. Extensive experiments are conducted on benchmark data sets to verify the effectiveness and the superiority of SUM-DCA in terms of F-score and rank-based evaluation without any bells and whistles.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑