Search papers, labs, and topics across Lattice.
1
0
3
0
Even when visual data is missing or noisy, EgoAdapt accurately determines who is talking to the camera wearer by adaptively integrating head orientation, lip movement, and robust audio features.
