In human action recognition, visual feature descriptors are of vital importance. based on RPCA,we get sparse component and low-rank of each video. Then we use the two-stream architecture.
Geek Repo:Geek Repo
Github PK Tool:Github PK Tool
