1

Find the Cliffhanger: Multi-modal Trailerness in Soap Operas
Predicting which moments in a video are trailer-worthy is a challenging task that benefits from multi-modal information.
Carlo Bretti, Pascal Mettes, Daan Odijk, Hendrik Vincent Koops, Nanne Van Noord
PDF Cite Code
Find the Cliffhanger: Multi-modal Trailerness in Soap Operas
Zero-Shot Action Recognition from Diverse Object-Scene Compositions
Predicting novel actions in videos without ever training on videos, only using objects and scenes.
Carlo Bretti, Pascal Mettes
PDF Cite Code
Zero-Shot Action Recognition from Diverse Object-Scene Compositions