search

KaiyangZhou / pytorch-vsumm-reinforce

Unsupervised video summarization with deep reinforcement learning (AAAI'18)
MIT License
464 stars 150 forks source link

Unsupervised learning use my own data #13

Open MuziSakura opened 5 years ago

MuziSakura commented 5 years ago

If I want to use my own dataset, but there is no features or labels in the dataset, only some video clips. How to extract features before construct the h5 file, and what should I do with user_summary, gts_score and gtsummary when construct h5 file?

Datartic commented 5 years ago

@MuziSakura were you able to do it on your own dataset?

cipher009 commented 4 years ago

Did anyone managed to test on own data?

Datartic commented 4 years ago

@cipher009 i managed to did it on my own data but the result was more like a video which was fast forwaded. It did manage to convert a 10 min long video to just 30 seconds.

This link might help you: https://medium.com/@shrutijadon10104776/video-summarization-an-survey-of-existing-algorithms-10b92a53e0a8

cipher009 commented 4 years ago

Thank you @syedimad1998, I will take a look. Which of the 5 methods did you use?

Datartic commented 4 years ago

@cipher009 4th worked for me pretty well, also look on the first and the last too. Feel free to mail me if there are any issues.

Tramac commented 4 years ago

@cipher009 4th worked for me pretty well, also look on the first and the last too. Feel free to mail me if there are any issues.

@syedimad1998 Hi, is it convenient to ask you some questions?

anaghazachariah commented 3 years ago

I implemented the project.We are using user summary,gtsummary and gtscore just to evaluate the model.You can develop the model in a fully unsupervised way. I implemented this project and you can take a look https://github.com/anaghazachariah/video_summary_generaton