Qwen-Audio-3.0-ASR Improves Entity Recall With Hierarchical Hotwords
An instruction-controlled MoE ASR model trained on tens of millions of hours reaches 99.43% recall for priority person names under hotword conditioning.
Sep 12, 20264 min2609.07549
Jianwei Yu
4 articles on SOTA Papers
Streaming ASR Unifies Transcription and Speaker Attribution
VibeVoice-ASR-Streaming interleaves audio, lookahead, and prior text so one model can produce speaker-attributed transcripts as speech arrives.
Sep 4, 20263 min2609.02812
Full-Context Rendering Reduces Music Codec Exposure Bias
FullDiT conditions a diffusion transformer on imperfect eight-stream codec plans, lyrics, and captions, improving ViSQOL by 0.77 under synthetic corruption.
Aug 12, 20264 min2608.08787
GigaSpeechBench Exposes ASR Gaps Beyond Standard Benchmarks
A 680-hour human-annotated benchmark combines multilingual, dialect, accent, domain, and age stress tests across ASR and speech translation.
Jul 28, 20265 min2606.28884