Bi-Directional Self-Attention with Relative Positional Encoding for Video Summarization.

Jingxu Lin, Sheng-hua Zhong

Browse the full ICTAI paper archive.