Citrinet: Closing the Gap between Non-Autoregressive and Autoregressive End-to-End Models for Automatic Speech Recognition
arXiv:2104.01721
Abstract
We propose Citrinet - a new end-to-end convolutional Connectionist Temporal Classification (CTC) based automatic speech recognition (ASR) model. Citrinet is deep residual neural model which uses 1D time-channel separable convolutions combined with sub-word encoding and squeeze-and-excitation. The resulting architecture significantly reduces the gap between non-autoregressive and sequence-to-sequence and transducer models. We evaluate Citrinet on LibriSpeech, TED-LIUM2, AISHELL-1 and Multilingual LibriSpeech (MLS) English speech datasets. Citrinet accuracy on these datasets is close to the best autoregressive Transducer models.
References in corpus (16)
- SpecAugment: A Simple Data Augmentation Method for Automatic Speech Recognition
- HuggingFace's Transformers: State-of-the-art Natural Language Processing
- Deep Speech 2: End-to-End Speech Recognition in English and Mandarin
- Sequence Transduction with Recurrent Neural Networks
- A Comparative Study on Transformer vs RNN in Speech Applications
- End-to-end Continuous Speech Recognition using Attention-based Recurrent NN: First Results
- Conformer: Convolution-augmented Transformer for Speech Recognition
- MLS: A Large-Scale Multilingual Dataset for Speech Research
- Listen, Attend and Spell
- NeMo: a toolkit for building AI applications using Neural Modules
- End-to-end ASR: from Supervised to Semi-Supervised Learning with Modern Architectures
- Stochastic Gradient Methods with Layer-wise Adaptive Moments for Training of Deep Networks
- The CAPIO 2017 Conversational Speech Recognition System
- ContextNet: Improving Convolutional Neural Networks for Automatic Speech Recognition with Global Context
- Unified Streaming and Non-streaming Two-pass End-to-end Model for Speech Recognition
- Recent Developments on ESPnet Toolkit Boosted by Conformer