mlsanityData to 7 Oct 2026

A Comparative Study on Neural Architectures and Training Methods for Japanese Speech Recognition

arXiv 2106.05111 ↗First posted 9 Jun 2021cs.CL cs.SD eess.AScited by 6

Not on the map. Why?

Abstract (arXiv)

End-to-end (E2E) modeling is advantageous for automatic speech recognition (ASR) especially for Japanese since word-based tokenization of Japanese is not trivial, and E2E modeling is able to model character sequences directly. This paper focuses on the latest E2E modeling techniques, and investigates their performances on character-based Japanese ASR by conducting comparative experiments. The results are analyzed and discussed in order to understand the relative advantages of long short-term memory (LSTM), and Conformer models in combination with connectionist temporal classification, transducer, and attention-based loss functions. Furthermore, the paper investigates on effectivity of the recent training techniques such as data augmentation (SpecAugment), variational noise injection, and exponential moving average. The best configuration found in the paper achieved the state-of-the-art character error rates of 4.1%, 3.2%, and 3.5% for Corpus of Spontaneous Japanese (CSJ) eval1, eval2, and eval3 tasks, respectively. The system is also shown to be computationally efficient thanks to the efficiency of Conformer transducers.

Key facts. Cited by 6 papers in the map (data through 7 Oct 2026). Builds on 19 of its 29 references that are in the map. First posted on arXiv 9 Jun 2021.

Builds on

19 of 29 references.

References · 19

  1. 2017Attention Is All You Needcited by 59,932
  2. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  3. 2014Neural Machine Translation by Jointly Learning to Align and Translatecited by 4,530
  4. 2020Conformer: Convolution-augmented Transformer for Speech Recognitioncited by 1,838
  5. 2019Transformer-XL: Attentive Language Models Beyond a Fixed-Length Contextcited by 1,830
  6. 2019SpecAugment: A Simple Data Augmentation Method for Automatic Speech Recognitioncited by 1,355
  7. 2016Language Modeling with Gated Convolutional Networkscited by 688
  8. 2018ESPnet: End-to-End Speech Processing Toolkitcited by 554
  9. 2019Libri-Light: A Benchmark for ASR with Limited or No Supervisioncited by 462
  10. 2019A Comparative Study on Transformer vs RNN in Speech Applicationscited by 210
  11. 2020Pushing the Limits of Semi-Supervised Learning for Automatic Speech Recognitioncited by 185
  12. 2020Recent Developments on ESPnet Toolkit Boosted by Conformercited by 132
7 more
  1. 2016Towards better decoding and language model integration in sequence to sequence modelscited by 100
  2. 2019Lingvo: a Modular and Scalable Framework for Sequence-to-Sequence Modelingcited by 86
  3. 2014End-to-end Continuous Speech Recognition using Attention-based Recurrent NN: First Resultscited by 68
  4. 2020On the Comparison of Popular End-to-End Models for Large Scale Speech Recognitioncited by 62
  5. 2017Advances in Joint CTC-Attention based End-to-End Speech Recognition with a Deep CNN Encoder and RNN-LMcited by 43
  6. 2019Scale MLPerf-0.6 models on Google TPU-v3 Podscited by 14
  7. 2016Optimizing Performance of Recurrent Neural Networks on GPUscited by 12

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 2; 2023: 3; 2024: 0; 2025: 0; 2026: 1.

6 papers.

  1. 2023Deep Model Fusion: A Surveycited by 89
  2. 2022Revisiting Checkpoint Averaging for Neural Machine Translationcited by 13
  3. 2022Structured State Space Decoder for Speech Recognition and Synthesiscited by 11
  4. 2023Retraining-free Customized ASR for Enharmonic Words Based on a Named-Entity-Aware Model and Phoneme Similarity Estimationcited by 5
  5. 2023Exploration of Language Dependency for Japanese Self-Supervised Speech Representation Modelscited by 3
  6. 2026Ruby-ASR: Evidence-Preserving Supervision for Joint Orthographic and Lexical-Reading Recognitioncited by 0