mlsanityData to 7 Oct 2026

Efficient Training of Neural Transducer for Speech Recognition

arXiv 2204.10586 ↗First posted 22 Apr 2022cs.CL cs.SD eess.AScited by 14

Not on the map. Why?

Abstract (arXiv)

As one of the most popular sequence-to-sequence modeling approaches for speech recognition, the RNN-Transducer has achieved evolving performance with more and more sophisticated neural network models of growing size and increasing training epochs. While strong computation resources seem to be the prerequisite of training superior models, we try to overcome it by carefully designing a more efficient training pipeline. In this work, we propose an efficient 3-stage progressive training pipeline to build highly-performing neural transducer models from scratch with very limited computation resources in a reasonable short time period. The effectiveness of each stage is experimentally verified on both Librispeech and Switchboard corpora. The proposed pipeline is able to train transducer models approaching state-of-the-art performance with a single GPU in just 2-3 weeks. Our best conformer transducer achieves 4.1% WER on Librispeech test-other with only 35 epochs of training.

Key facts. Cited by 14 papers in the map (data through 7 Oct 2026). Builds on 28 of its 45 references that are in the map. First posted on arXiv 22 Apr 2022.

Builds on

28 of 45 references.

References · 28

  1. 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,267
  2. 2015Rethinking the Inception Architecture for Computer Visioncited by 5,451
  3. 2020Conformer: Convolution-augmented Transformer for Speech Recognitioncited by 1,838
  4. 2015Listen, Attend and Spellcited by 542
  5. 2020Transformer Transducer: A Streamable Speech Recognition Model with Transformer Encoders and RNN-T Losscited by 232
  6. 2015On Using Monolingual Corpora in Neural Machine Translationcited by 149
  7. 2015End-to-End Attention-based Large Vocabulary Speech Recognitioncited by 148
  8. 2020Hybrid Autoregressive Transducer (hat)cited by 106
  9. 2019Language Modeling with Deep Transformerscited by 96
  10. 2020ContextNet: Improving Convolutional Neural Networks for Automatic Speech Recognition with Global Contextcited by 96
  11. 2019Transformer-based Acoustic Modeling for Hybrid Speech Recognitioncited by 86
  12. 2020Internal Language Model Estimation for Domain-Adaptive End-to-End Speech Recognitioncited by 80
16 more
  1. 2020A Density Ratio Approach to Language Model Fusion in End-To-End Automatic Speech Recognitioncited by 73
  2. 2019SpecAugment on Large Scale Datasetscited by 60
  3. 2021Advancing RNN Transducer Technology for Speech Recognitioncited by 44
  4. 2020Single headed attention based sequence-to-sequence model for state-of-the-art results on Switchboardcited by 36
  5. 2020A New Training Pipeline for an Improved Neural Transducercited by 34
  6. 2020Efficient minimum word error rate training of RNN-Transducer for end-to-end speech recognitioncited by 34
  7. 2021On the limit of English conversational speech recognitioncited by 30
  8. 2020Less Is More: Improved RNN-T Decoding Using Limited Label Context and Path Mergingcited by 26
  9. 2021Investigating Methods to Improve Language Model Integration for Attention-based Encoder-Decoder ASR Modelscited by 25
  10. 2019Minimum Bayes Risk Training of RNN-Transducer for End-to-End Speech Recognitioncited by 23
  11. 2020Phoneme Based Neural Transducer for Large Vocabulary Speech Recognitioncited by 20
  12. 2020The RWTH ASR System for TED-LIUM Release 2: Improving Hybrid HMM with SpecAugmentcited by 19
  13. 2021Minimum Word Error Rate Training with Language Model Fusion for End-to-End Speech Recognitioncited by 19
  14. 2021On Language Model Integration for RNN Transducer based Speech Recognitioncited by 17
  15. 2021Conformer-based Hybrid ASR System for Switchboard Datasetcited by 15
  16. 2021Acoustic Data-Driven Subword Modeling for End-to-End Speech Recognitioncited by 11

Built on by

Cited by year

Citing papers per year: 2022: 2; 2023: 8; 2024: 1; 2025: 3.

14 papers.

Sort:

  1. 2023End-to-End Speech Recognition: A Surveycited by 127
  2. 2024Dynamic Encoder Size Based on Data-Driven Layer-wise Pruning for Speech Recognitioncited by 7
  3. 2022Enhancing and Adversarial: Improve ASR with Speaker Labelscited by 6
  4. 2023Towards Effective and Compact Contextual Representation for Conformer Transducer Speech Recognition Systemscited by 6
  5. 2023Comparative Analysis of the wav2vec 2.0 Feature Extractorcited by 4
  6. 2022Lattice-Free Sequence Discriminative Training for Phoneme-Based Neural Transducerscited by 3
  7. 2023RASR2: The RWTH ASR Toolkit for Generic Sequence-to-sequence Speech Recognitioncited by 3
  8. 2023On the Relation between Internal Language Model and Sequence Discriminative Training for Neural Transducerscited by 1
  9. 2023End-to-End Training of a Neural HMM with Label and Transition Probabilitiescited by 1
  10. 2025Regularizing Learnable Feature Extraction for Automatic Speech Recognitioncited by 1
  11. 2025Dynamic Acoustic Model Architecture Optimization in Training for ASRcited by 1
  12. 2023Analyzing And Improving Neural Speaker Embeddings for ASRcited by 0
2 more
  1. 2023Investigating the Effect of Language Models in Sequence Discriminative Training for Neural Transducerscited by 0
  2. 2025Unified Learnable 2D Convolutional Feature Extraction for ASRcited by 0