Search papers, labs, and topics across Lattice.
5
0
5
DELTA-TTS achieves 3.3x faster token generation while reducing word error rates, transforming the landscape of text-to-speech synthesis.
Bridging the gap between context awareness and adherence could redefine how spoken dialogue systems maintain coherence across conversations.
NaturalFlow reduces disruptive pauses in simultaneous translation, achieving smoother speech flow without sacrificing latency or translation quality.
Shifting from token-level to patch-level modeling in TTS can yield a 1.8x speedup and drastically cut memory usage.
Voice assistants often fail to distinguish between a primary user and a third-party interruption, but this new dataset and benchmark could fix that.
