Unsupervised Uncertainty Measures of Automatic Speech Recognition for Non-intrusive Speech Intelligibility Prediction
Unsupervised Uncertainty Measures of Automatic Speech Recognition for Non-intrusive Speech Intelligibility Prediction
复制标题
用于非侵入式语音清晰度预测的自动语音识别的无监督不确定性测量
DOI：
--
复制
发表时间：
2022
期刊：
影响因子：
--
通讯作者：
J. Barker
中科院分区：
文献类型：
--
作者：
Zehai Tu;Ning Ma;J. Barker
非侵入式可懂度预测对于其在难以获得干净参考信号的现实场景中的应用是重要的。许多非侵入式预测器的构建需要地面真实可理解性标签或用于监督学习的干净参考信号。在这项工作中，我们利用一种无监督的不确定性估计方法来预测语音清晰度，该方法不需要清晰度标签或参考信号来训练预测器。我们的实验表明，从国家的最先进的端到端的自动语音识别（ASR）模型的不确定性是高度相关的语音清晰度。该方法在两个数据库上进行了评估，结果表明，ASR模型的无监督不确定性措施与听力结果的语音可懂度比广泛使用的侵入性方法的预测更相关。
Non-intrusive intelligibility prediction is important for its application in realistic scenarios, where a clean reference signal is difficult to access. The construction of many non-intrusive predictors require either ground truth intelligibility labels or clean reference signals for supervised learning. In this work, we leverage an unsupervised uncertainty estimation method for predicting speech intelligibility, which does not require intelligibility labels or reference signals to train the predictor. Our experiments demonstrate that the uncertainty from state-of-the-art end-to-end automatic speech recognition (ASR) models is highly correlated with speech intelligibility. The proposed method is evaluated on two databases and the results show that the unsupervised uncertainty measures of ASR models are more correlated with speech intelligibility from listening results than the predictions made by widely used intrusive methods.
第一届清晰度预测挑战赛：助听器清晰度预测的机器学习挑战赛
DOI：
10.21437/interspeech.2022-10821
发表时间：
2022
期刊：
--
影响因子：
--
作者：
Barker J
通讯作者：
Barker J
使用深度神经网络预测语音清晰度
DOI：
10.1016/j.csl.2017.10.004
发表时间：
2018
期刊：
Comput. Speech Lang.
影响因子：
--
作者：
Spille C;Kollmeier B;Meyer BT
通讯作者：
Meyer BT