Learning-Action-Completeness-from-Points
Learning-Action-Completeness-from-Points copied to clipboard
Pilhyeon
about feature extraction of THUMOS14
Excuse me, what methuod do you use to extract the feature of thumos14? Is it I3D?
Yes, we used the two-stream I3D networks pre-trained on Kinetics-400, as stated in the paper.
Yes, we used the two-stream I3D networks pre-trained on Kinetics-400, as stated in the paper.
ok，thank you.You said in your paper that you take 16 frames of video as a segment, and then extract features. But when I output the shape (nxdimension) of the feature, n is not equal to the total number of frames /16. Why is this?
It is probably because we set the fps of all videos in THUMOS'14 to 25 before extracting the features.
Yes，that makes sense！Thank you very much！
Hello，I am a PhD student from China. I am also interested in this work now. I have some issues about this work. Can we make a discussion? My WeChat ID is fj_snbc.
