Hearing a face: cross-modal speaker matching using isolated visible speech.

Hearing a face: cross-modal speaker matching using isolated visible speech.