Skip to content
PyTorch Implementation of SUM-GAN from "Unsupervised Video Summarization with Adversarial LSTM Networks" (CVPR 2017)
Branch: master
Clone or download
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
imgs Update readme: figures Nov 15, 2017
layers Fix KL-divergence, plot verbosely Nov 16, 2017
README.md Update Readme: googlenet pool5 Nov 16, 2017
configs.py Fix KL-divergence, plot verbosely Nov 16, 2017
data_loader.py Change evaluation scheme Nov 15, 2017
feature_extraction.py Refactor Nov 15, 2017
solver.py Fix KL-divergence, plot verbosely Nov 16, 2017
train.py Change evaluation scheme Nov 15, 2017
utils.py Refactor Nov 15, 2017

README.md

Adversarial_Video_Summary

PyTorch Implementation of SUM-GAN

Changes from Original paper

  • Video feature extractor
    • GoogleNet pool5 (1024) => ResNet-101 pool5 (2048)
    • Followed by linear projection to 500-dim
  • Stable GAN Training
    • Discriminator's learning rate: 1e-5 (Others: 1e-4)
    • Fix Discriminators' parameters for first 15 steps at every epoch.

Model figures

Model figure 1

Model figure 2

Model figure 3

Algorithm

Algorithm figure

You can’t perform that action at this time.