Search papers, labs, and topics across Lattice.
4
0
6
A novel reward system boosts Qwen-Image-2.0's performance, achieving a 2.61 point increase in overall quality and significant gains in both text-to-image and image editing tasks.
Achieving high accuracy in multi-speaker transcription, SoulX-Transcriber outperforms existing models by effectively addressing speaker overlap and rapid turn-taking.
LLMs can learn musicality without human annotation by aligning them to automatically generated preference datasets derived from rule-based musical constraints.
Achieve human-like full-duplex voice interactions with SoulX-Duplug, a plug-and-play module that slashes latency and improves turn management by acting as a semantic VAD.
