search

facebookarchive / C3D

C3D is a modified version of BVLC caffe to support 3D ConvNets.
Other
1.16k stars 506 forks source link

Extracting features for new videos #426

Closed kjsr7 closed 5 years ago

kjsr7 commented 5 years ago

I have successfully extracted the features for the example videos using the script c3d_sport1m_feature_extraction_video.sh. I have learned from the guide on how to extract features from new videos. But how do I know the number of frames of new videos so that I can specify them in the input list?

kjsr7 commented 5 years ago

You can use this

Shiny128 commented 5 years ago

Hello Sir, Could you please describe the steps to be carried out to extract features for new set of videos?

Thank You