Quartznet: Deep Automatic Speech Recognition with 1D Time-Channel Separable Convolutions
Samuel Kriman, Stanislav Beliaev, Boris Ginsburg, Jocelyn Huang, Oleksii Kuchaiev, Vitaly Lavrukhin, R. Bret Leary, Jason Li, Yang Zhang
Abstract
We propose a new end-to-end neural acoustic model for automatic speech recognition. The model is composed of multiple blocks with residual connections between them. Each block consists of one or more modules with 1D time-channel separable convolutional layers, batch normalization, and ReLU layers. It is trained with CTC loss. The proposed network achieves near state-of-the-art accuracy on LibriSpeech and Wall Street Journal, while having fewer parameters than all competing models. We also demonstrate that this model can be effectively fine-tuned on new datasets.
Topics & Concepts
Normalization (sociology)Computer scienceSpeech recognitionConvolutional neural networkSeparable spaceResidualChannel (broadcasting)Block (permutation group theory)Acoustic modelConvolution (computer science)Pattern recognition (psychology)State (computer science)Artificial intelligenceAlgorithmArtificial neural networkSpeech processingMathematicsTelecommunicationsMathematical analysisGeometrySociologyAnthropologySpeech Recognition and SynthesisSpeech and Audio ProcessingMusic and Audio Processing