Search papers, labs, and topics across Lattice.
1
0
2
8
Interleaving speech and text during ASR training boosts entity recognition accuracy and narrows the gap between modalities, challenging traditional training paradigms.
