Search papers, labs, and topics across Lattice.
1
0
2
6
Filtering speech data with an audio-language model boosts translation accuracy by over 1.4 ASR-BLEU, highlighting the critical role of data quality in S2ST.