研究人员提出了一种用于评估自动语音识别（ASR）系统的新范式，旨在改进现有的词错误率（WER）和字符错误率（CER）等指标。所提出的方法结合了选定的指标来生成最小编辑距离（minED），该距离与人类感知有更好的相关性，并考虑了语言和语义信息。这种方法允许从人类的角度更细致地研究转录错误的严重性。 AI
影响 这种新的评估范式可能带来更准确、更符合人类习惯的ASR系统，从而影响依赖语音转录的下游应用。
排序理由 该集群包含一篇详细介绍ASR评估新方法的学术论文。
AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →
arXiv:2605.03671v1 Announce Type: new Abstract: The most commonly used metrics for evaluating automatic speech transcriptions, namely Word Error Rate (WER) and Character Error Rate (CER), have been heavily criticized for their poor correlation to human perception and their inabil…
The most commonly used metrics for evaluating automatic speech transcriptions, namely Word Error Rate (WER) and Character Error Rate (CER), have been heavily criticized for their poor correlation to human perception and their inability to take into account linguistic and semantic…