Learning speaker, addressee and overlap detection models from multimodal streams.

Oriol Vinyals, Dan Bohus, Rich Caruana

Browse the full ICMI paper archive.