PulseAugur
实时 07:28:36
English(EN) RAS: a Reliability Oriented Metric for Automatic Speech Recognition

研究人员推出RAS,一种用于可靠语音识别系统的新度量指标

研究人员推出了一种名为RAS的新度量指标,旨在评估自动语音识别(ASR)系统的可靠性。与仅关注准确性的传统度量指标不同,RAS考虑了系统在转录中的置信度,尤其是在嘈杂或模糊的条件下。所提出的框架允许ASR模型在不确定的片段上弃权,RAS通过人类偏好进行校准,平衡了转录的信息量和避免错误。实验表明,这种方法在保持具有竞争力的准确性的同时,显著提高了转录的可靠性。 AI

影响 为ASR系统引入了一种新的评估度量指标,该指标在准确性之外优先考虑可靠性,有可能提高用户信任度和下游应用程序的性能。

排序理由 介绍ASR可靠性新度量指标的学术论文。

在 arXiv cs.AI 阅读 →

AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →

研究人员推出RAS,一种用于可靠语音识别系统的新度量指标

报道来源 [2]

  1. arXiv cs.AI TIER_1 English(EN) · Wenbin Huang, Yuhang Qiu, Bohan Li, Yiwei Guo, Jing Peng, Hankun Wang, Xie Chen, Kai Yu ·

    RAS: a Reliability Oriented Metric for Automatic Speech Recognition

    arXiv:2604.24278v1 Announce Type: cross Abstract: Automatic speech recognition systems often produce confident yet incorrect transcriptions under noisy or ambiguous conditions, which can be misleading for both users and downstream applications. Standard evaluation based on Word E…

  2. arXiv cs.AI TIER_1 English(EN) · Kai Yu ·

    RAS: a Reliability Oriented Metric for Automatic Speech Recognition

    Automatic speech recognition systems often produce confident yet incorrect transcriptions under noisy or ambiguous conditions, which can be misleading for both users and downstream applications. Standard evaluation based on Word Error Rate focuses solely on accuracy and fails to …