发表机构
Stimuler（Stimuler）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
针对带口音对话式ASR，提出三阶段流水线，通过数据策展和区域LoRA适配器，将实体召回率提升至80-85%，填充词召回率提升至76-86%，并保持低WER。
AI 中文摘要
针对词错误率（WER）优化的ASR系统常常遗漏带口音对话式英语中的命名实体和填充停顿，而这两者对于语言学习反馈至关重要。我们针对来自印度、印度尼西亚和拉丁美洲的说话者提出了一种三阶段流水线：（1）启发式SQL过滤器，以随机采样2.8倍的实体密度策展富含实体的训练数据；（2）在Qwen2.5-Omni-3B上微调的区域LoRA适配器，在单次前向传播中同时生成逐字和纠正后的转录；（3）由基于LLM的评判者验证的六类别错误分类法（83.8%一致性，210个人工标注样本）。该流水线在6k测试话语上实现了80-85%的实体召回率（从53-55%提升）、76-86%的填充词召回率（从<5%提升）和6-10%的WER，在实体召回率上优于Whisper和商业ASR，同时以少10倍的参数匹配零样本30B模型。配对自助法检验确认，仅策展就贡献了2.8-4.2个百分点的实体召回率提升（p<0.0001）。
英文摘要
ASR systems optimised for Word Error Rate (WER) often miss named entities and filled pauses in accented conversational English, both critical for language-learning feedback. We present a three-stage pipeline for speakers from India, Indonesia, and Latin America: (1) heuristic SQL filters curating entity-rich training data at 2.8x the entity density of random sampling, (2) regional LoRA adapters fine-tuned on Qwen2.5-Omni-3B producing both verbatim and corrected transcripts in a single forward pass, and (3) a six-category error taxonomy validated by an LLM-based judge (83.8% agreement, 210 human-labelled samples). The pipeline achieves 80-85% entity recall (up from 53-55%), 76-86% filler recall (up from <5%), and 6-10% WER across 6k test utterances, outperforming Whisper and a commercial ASR on entity recall while matching a zero-shot 30B model with 10x fewer parameters. Paired bootstrap tests confirm that curation alone accounts for 2.8-4.2 pp of entity recall gain (p<0.0001).
Comments5 pages, 1 figure, 1 table, accepted at Interspeech 2026
