1 paper · 1 filter
Fanqi Yu, Shengming Ma, Stefano Fiorini +4
Fully supervised repetitive action counting (RAC) has achieved strong performance, but requires dense temporal annotations that are costly and difficult to scale. We propose TReViS…