1 paper · 1 filter
Barak Battash, Haim Barad, Hanlin Tang +1
Video understanding usually requires expensive computation that prohibits its deployment, yet videos contain significant spatiotemporal redundancy that can be exploited. In particu…