weirme

I think each frame corresponds to a score in the tsv file. Assuming a frame rate of 30 frames per sec, the 1st video with a length of 5 min...

Thank you a lot, I think you are right and I have fixed this problem. But it's strange that F-score after modification seems to be almost same as before...

> > Thank you a lot, I think you are right and I have fixed this problem. But it's strange that F-score after modification seems to be almost same as...

> I implement myself by revising your code. > I got the performance quite well. > [github](https://github.com/pcshih/pytorch-FCSN/tree/master) Yes, this code performs well on TVSum dataset.

I think you are right, my original code convert multi-user F-score to a single... We may make the dataset using a similar method in evaluation.

This part in paper is a bit confusing, but I think the result of your implantation is good!

> Could you give me the Summe dataset (downsampled to 320 frames per video) preprocessed by make_dataset.py? > I would like to try the evaluation. Sorry, the hdf5 file is...

> Is the performance far from the paper mentioned(SumME)? Yes

> I cannot run the code. > What is the input format of video? > i.e. Air_Force_One.mp4 You may run `python make_dataset.py --video_dir {arg1} --h5_path {arg2} --vsumm_data {arg3}` and here...

Almost as bad as yours :(