CarneliNet: Neural Mixture Model for Automatic Speech Recognition
This addresses streaming ASR performance issues for real-time applications, but it is incremental as it builds on existing CTC-based methods with a novel architectural tweak.
The paper tackles the problem of deep end-to-end ASR models having large receptive fields that hurt streaming performance, by proposing CarneliNet, a CTC-based neural mixture model using parallel shallow networks, which achieves close to state-of-the-art results on LibriSpeech, MLS, and AISHELL-2 datasets and allows dynamic reconfiguration without retraining.
End-to-end automatic speech recognition systems have achieved great accuracy by using deeper and deeper models. However, the increased depth comes with a larger receptive field that can negatively impact model performance in streaming scenarios. We propose an alternative approach that we call Neural Mixture Model. The basic idea is to introduce a parallel mixture of shallow networks instead of a very deep network. To validate this idea we design CarneliNet -- a CTC-based neural network composed of three mega-blocks. Each mega-block consists of multiple parallel shallow sub-networks based on 1D depthwise-separable convolutions. We evaluate the model on LibriSpeech, MLS and AISHELL-2 datasets and achieved close to state-of-the-art results for CTC-based models. Finally, we demonstrate that one can dynamically reconfigure the number of parallel sub-networks to accommodate the computational requirements without retraining.