AuraTracer智迹闻
中文

EVENT DOSSIER

Automatic Speech Recognition for Multilingual Oral History Research

2026-09-07 12:00 Models 🔥 40.2 heat score
1sources
1days unfolding
40.2heat score
2mentions
SummaryAI generated

On September 7, 2026, the arXiv cs.CL platform released an automatic speech recognition (ASR) system for the study of multilingual oral history. This system aims to address the challenges posed by language diversity in oral history research, achieving accurate recognition and processing of spoken content in multiple languages through technical means, providing new tool support for academic research in this field.

Related eventsRELATED EVENTS
Key entitiesKEY ENTITIES
WhisperarXiv

Coverage · reports per dayLANGUAGE SPLIT

Entity relations
Whisper × arXiv1

SignalsSIGNALS

Keyword heat
  • Whisper1
  • arXiv1

All reports (1)SOURCES

A arXiv cs.CL en 2026-09-07 12:00

Automatic Speech Recognition for Multilingual Oral History Research

本文提出社区主导的文化遗产语言保护与复兴项目正采用语音技术。在新西兰粤语复兴中,口述历史作为关键的语言维护策略,其转录工作常因耗时耗力而受阻。自动语音识别(ASR)工具包如 Whisper 的开发加速了这一过程,但针对语码转换语境的有效性研究有限。基于字错率(WER),表现最佳的 Whisper 模型配置 WER 为 12.10,代价是未能准确转录非英语部分;尽管如此,该工具仍有用,仅需占手动转录预估时间的 1% 即可提供初稿转录。