SUM-FCN

A Pytorch implementation of Video Summarization Using Fully Convolutional Sequence Networks

Paper Publication Info

Mrigank Rochan, Linwei Ye, and Yang Wang.
Video Summarization Using Fully Convolutional Networks.
European Conference on Computer Vision (ECCV), 2018

pdf link: http://openaccess.thecvf.com/content_ECCV_2018/papers/Mrigank_Rochan_Video_Summarization_Using_ECCV_2018_paper.pdf

this implementation uses Sigmoid+BCE instead of Softmax+CE