Robust Raw Waveform Speech Recognition Using Relevance Weighted Representations
This work addresses noise robustness in speech recognition for applications in real-world environments, representing an incremental advancement with specific gains.
The paper tackled the problem of speech recognition in noisy and channel-distorted scenarios by developing a novel acoustic modeling framework using relevance weighting for feature selection, resulting in significant improvements with an average relative reduction of 10% in word error rates over baseline systems on multiple datasets.
Speech recognition in noisy and channel distorted scenarios is often challenging as the current acoustic modeling schemes are not adaptive to the changes in the signal distribution in the presence of noise. In this work, we develop a novel acoustic modeling framework for noise robust speech recognition based on relevance weighting mechanism. The relevance weighting is achieved using a sub-network approach that performs feature selection. A relevance sub-network is applied on the output of first layer of a convolutional network model operating on raw speech signals while a second relevance sub-network is applied on the second convolutional layer output. The relevance weights for the first layer correspond to an acoustic filterbank selection while the relevance weights in the second layer perform modulation filter selection. The model is trained for a speech recognition task on noisy and reverberant speech. The speech recognition experiments on multiple datasets (Aurora-4, CHiME-3, VOiCES) reveal that the incorporation of relevance weighting in the neural network architecture improves the speech recognition word error rates significantly (average relative improvements of 10% over the baseline systems)