Gang Li

2 articles on SOTA Papers

End-to-End Audio Generation Cuts Speech Errors

An LLM predicts audio latents autoregressively while per-token flow matching generates variable-length scenes, reducing English Seed-TTS WER from 12.15% to 2.79%.

Aug 25, 20264 min2608.11804 Code available

FPGA Control Cuts Atom Rearrangement Latency

A PC-free PXIe feedback loop integrates counting, path planning, and waveform control, reaching 282 μs latency and 95.4% 10-atom assembly success.

Jul 12, 20265 min2607.08687