Predicting word error rate for reverberant speech
This addresses the problem of quantifying ASR performance degradation due to reverberation for speech recognition systems, with incremental improvements in prediction methods.
The paper tackled predicting word error rate (WER) for reverberant speech in automatic speech recognition by proposing methods to estimate WER from acoustic parameters or blindly from speech samples, finding that clarity parameters (C50, C80) are highly correlated with WER and a CNN model outperforms parameter-based prediction.
Reverberation negatively impacts the performance of automatic speech recognition (ASR). Prior work on quantifying the effect of reverberation has shown that clarity (C50), a parameter that can be estimated from the acoustic impulse response, is correlated with ASR performance. In this paper we propose predicting ASR performance in terms of the word error rate (WER) directly from acoustic parameters via a polynomial, sigmoidal, or neural network fit, as well as blindly from reverberant speech samples using a convolutional neural network (CNN). We carry out experiments on two state-of-the-art ASR models and a large set of acoustic impulse responses (AIRs). The results confirm C50 and C80 to be highly correlated with WER, allowing WER to be predicted with the proposed fitting approaches. The proposed non-intrusive CNN model outperforms C50-based WER prediction, indicating that WER can be estimated blindly, i.e., directly from the reverberant speech samples without knowledge of the acoustic parameters.