shuyueL / FFNet

Implementation of our CVPR 2018 Paper (FFNet: Video Fast-Forwarding via Reinforcement Learning)
GNU General Public License v3.0
20 stars 4 forks source link

Thank for Your works #3

Closed zmxzsl closed 3 years ago

zmxzsl commented 5 years ago

I want to know whether the data required for training needs to be extract feature by network up front,saving file .mat?Thanks again.

shuyueL commented 5 years ago

Yes. the input of FFNet is a feature vector extracted from Alexnet.

nikhileshva commented 5 years ago

Also, is the feature vector for a single video file or for the entire dataset?

shuyueL commented 5 years ago

@nikhileshva The input is a feature vector of dimension 4096, which is extracted with a single frame.