Video-Summarization-with-LSTM
                                
                                 Video-Summarization-with-LSTM copied to clipboard
                                
                                    Video-Summarization-with-LSTM copied to clipboard
                            
                            
                            
                        Intersection over Union
Why is user intersection used for evaluating SumMe and not done in the same way as TVSum? I couldn't spot the reason why in the paper.
Thanks for making your code available!