On Architectures and Training for Raw Waveform Feature Extraction in ASR
This work addresses feature extraction for ASR systems, but it is incremental as it focuses on applying and combining existing methods rather than introducing new ones.
The study investigated the effectiveness of the wav2vec front-end framework for raw waveform feature extraction in hybrid automatic speech recognition (ASR) systems without additional untranscribed data, comparing it to Gammatone features and supervised acoustic model features, and exploring combinations to improve performance.
With the success of neural network based modeling in automatic speech recognition (ASR), many studies investigated acoustic modeling and learning of feature extractors directly based on the raw waveform. Recently, one line of research has focused on unsupervised pre-training of feature extractors on audio-only data to improve downstream ASR performance. In this work, we investigate the usefulness of one of these front-end frameworks, namely wav2vec, in a setting without additional untranscribed data for hybrid ASR systems. We compare this framework both to the manually defined standard Gammatone feature set, as well as to features extracted as part of the acoustic model of an ASR system trained supervised. We study the benefits of using the pre-trained feature extractor and explore how to additionally exploit an existing acoustic model trained with different features. Finally, we systematically examine combinations of the described features in order to further advance the performance.