Cross-Language Transfer Learning, Continuous Learning, and Domain Adaptation for End-to-End Automatic Speech Recognition
arXiv:2005.04290
Abstract
In this paper, we demonstrate the efficacy of transfer learning and continuous learning for various automatic speech recognition (ASR) tasks. We start with a pre-trained English ASR model and show that transfer learning can be effectively and easily performed on: (1) different English accents, (2) different languages (German, Spanish and Russian) and (3) application-specific domains. Our experiments demonstrate that in all three cases, transfer learning from a good base model has higher accuracy than a model trained from scratch. It is preferred to fine-tune large models than small pre-trained models, even if the dataset for fine-tuning is small. Moreover, transfer learning significantly speeds up convergence for both very small and very large target datasets.
References in corpus (5)
- Listen, Attend and Spell
- Common Voice: A Massively-Multilingual Speech Corpus
- NeMo: a toolkit for building AI applications using Neural Modules
- Stochastic Gradient Methods with Layer-wise Adaptive Moments for Training of Deep Networks
- Multilingual Training and Cross-lingual Adaptation on CTC-based Acoustic Model
Cited by in corpus (5)
- Adaptation Algorithms for Neural Network-Based Speech Recognition: An Overview
- SPGISpeech: 5,000 hours of transcribed financial audio for fully formatted end-to-end speech recognition
- A Simple Baseline for Domain Adaptation in End to End ASR Systems Using Synthetic Data
- E2E-based Multi-task Learning Approach to Joint Speech and Accent Recognition
- Coarse-To-Fine And Cross-Lingual ASR Transfer