1 paper
Anurag Bagchi, Jazib Mahmood, Dolton Fernandes +1
State of the art architectures for untrimmed video Temporal Action Localization (TAL) have only considered RGB and Flow modalities, leaving the information-rich audio modality tota…