Feb 32021 Experimental AI framework Vx2Text generates video captions using inferences from audio and text Researchers at Facebook have developed a framework, Vx2Text, that can generate captions by inferring from videos, audio, and text.
