Exploring Global Diversity and Local Context for Video Summarization
This work addresses video summarization for large-scale video processing, offering an incremental improvement over existing methods.
The paper tackled the problem of generating diverse video summaries by identifying that self-attention mechanisms using inner-product affinity produce discriminative rather than diversified features, and proposed a model combining global diverse attention with squared Euclidean distance and local contextual attention, achieving improved F-score and rank-based evaluations on benchmark datasets.
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.