Search papers, labs, and topics across Lattice.
1
0
3
By explicitly modeling speech, SAVE leapfrogs existing audio-visual methods for video-text retrieval, achieving substantial gains over the state-of-the-art.
