Speculative Listening speeds speech models using idle compute
It achieves 2.6-2.8x speedup on speak chunks and reduces delay by 73%.
Stories per day · 30 days
3 RSS
8 stories, newest first · one card per story
Stories tagged #speech-llm
Speech LLM ASR expansion forgets source languages unless merges are selective
SSU-Wanda+TIES best preserved source languages in single-target expansion; recursive multi-target updates were highly order-sensitive.
DirectSpeech2LLM: A Simple End-to-End Framework to Mitigate Prompt Overfitting in Speech-LLMs
DirectSpeech2LLM cuts speech-LLM prompt overfitting, generalizing from ASR-only training to translation and emotion tasks.
HINTT Submission to the 2nd MLC-SLM Challenge: Comparing Cascaded and Unified Approaches to Diarization and ASR
HINTT's cascaded diarization and Qwen3-ASR pipeline beat a unified speech LLM on MLC-SLM Task 1.
Pruning for Efficiency, Paying in Fairness: Demographic Disparities in Pruned Speech-LLMs
Pruning SLAM-ASR audio encoders widens demographic WER gaps that aggregate error rates hide.
Nuha-Speech: Building General-Purpose Arabic Speech-LLMs
Nuha-Speech initiative builds general-purpose Arabic speech-LLMs using a 1.5M-sample speech QA corpus and fine-tuned Qwen-Omni variants.
RetroThinker: Enabling Retrospective Thinking in Speech LLMs
RetroThinker is a post-training framework letting the Moshi speech LLM self-correct reasoning mid-stream, adding 11% GSM8K accuracy at similar latency.
X-AuT: Progressive Audio-Encoder Compression for Speech LLMs with Cross-Scale Distillation
XPeng AI's X-AuT prunes speech LLM audio encoders, cutting Qwen3-ASR-0.6B error from 5.61% to 5.27% with fewer parameters.