Video Violence Recognition and Localization Using a Semi-Supervised Hard Attention Model

Open in new window