一句话：实证研究揭示 TTS 合成语音测试用例在 ASR 测试中会产生误报，即合成音频失败但真人音频可正确转录。
近期研究提出利用文本到语音（TTS）系统自动合成大规模语音测试用例，以发现自动语音识别（ASR）系统中的大量失败。然而，合成测试用例所发现的失败可能并不反映 ASR 在转录真人音频时的真实性能，本文将这种现象称为误报。给定一个由 TTS 系统合成的失败测试用例，其中包含 TTS 生成的音频与对应的真值文本，作者将朗读同一文本的真人音频输入 ASR 系统；如果真人音频能被正确转录，则检测到一个误报实例。论文围绕该现象开展实证研究，旨在评估 TTS 合成测试用例在 ASR 测试中的有效性，并分析其误报问题。
Recent studies have proposed the use of Text-To-Speech (TTS) systems to automatically synthesise speech test cases on a scale and uncover a large number of failures in ASR systems. However, the failures uncovered by synthetic test cases may not reflect the actual performance of an ASR system when it transcribes human audio, which we refer to as false alarms. Given a failed test case synthesised from TTS systems, which consists of TTS-generated audio and the corresponding ground truth text, we feed the human audio stating the same text to an ASR system. If human audio can be correctly transcribed, an instance of a false alarm is detected.
@inproceedings{paperbot8056,
title = {Synthesizing Speech Test Cases with Text-to-Speech? An Empirical Study on the False Alarms in Automated Speech Recognition Testing},
author = {Julia Kaiwen Lau and Kelvin Kai Wen Kong and Julian Hao Yong and Per Hoong Tan and Zhou Yang and Zi Qian Yong and Joshua Chern Wey Low and Chun Yong Chong and Mei Kuan Lim and David Lo},
booktitle = {ISSTA},
year = {2023},
doi = {10.1145/3597926.3598126}
}