mlsanityData to 14 Oct 2026

Echo State Speech Recognition

arXiv 2102.09114 ↗First posted 18 Feb 2021cs.CLcited by 11

Not on the map. Why?

Abstract (arXiv)

We propose automatic speech recognition (ASR) models inspired by echo state network (ESN), in which a subset of recurrent neural networks (RNN) layers in the models are randomly initialized and untrained. Our study focuses on RNN-T and Conformer models, and we show that model quality does not drop even when the decoder is fully randomized. Furthermore, such models can be trained more efficiently as the decoders do not require to be updated. By contrast, randomizing encoders hurts model quality, indicating that optimizing encoders and learn proper representations for acoustic inputs are more vital for speech recognition. Overall, we challenge the common practice of training ASR models for all components, and demonstrate that ESN-based models can perform equally well but enable more efficient training and storage than fully-trainable counterparts.

Key facts. Cited by 11 papers in the map (data through 14 Oct 2026). Builds on 9 of its 23 references that are in the map. First posted on arXiv 18 Feb 2021.

Builds on

9 of 23 references.

References · 9

  1. 2017Attention Is All You Needcited by 60,361
  2. 2015Neural Machine Translation of Rare Words with Subword Unitscited by 2,918
  3. 2020Conformer: Convolution-augmented Transformer for Speech Recognitioncited by 1,863
  4. 2013Speech Recognition with Deep Recurrent Neural Networkscited by 669
  5. 2017State-of-the-art Speech Recognition With Sequence-to-Sequence Modelscited by 196
  6. 2020A Streaming On-Device End-to-End Model Surpassing Server-Side Conventional Model Quality and Latencycited by 100
  7. 2018Exploring Architectures, Data and Units For Streaming End-to-End Speech Recognition with RNN-Transducercited by 96
  8. 2018Comparison between DeepESNs and gated RNNs on multivariate time-series predictioncited by 4
  9. 2019Richness of Deep Echo State Network Dynamicscited by 3

Built on by

Cited by year

Citing papers per year: 2021: 5; 2022: 1; 2023: 5.

11 papers.

  1. 2023Efficient Sequence Transduction by Jointly Predicting Tokens and Durationscited by 58
  2. 2021Tied & Reduced RNN-T Decodercited by 36
  3. 2021Efficient and Private Federated Learning with Partially Trainable Networkscited by 13
  4. 2023CIF-T: A Novel CIF-based Transducer Architecture for Automatic Speech Recognitioncited by 6
  5. 2023Replay to Remember: Continual Layer-Specific Fine-tuning for German Speech Recognitioncited by 4
  6. 2023Neural Graph Revealerscited by 3
  7. 2021Compressing 1D Time-Channel Separable Convolutions using Sparse Random Ternary Matricescited by 2
  8. 2021Raw Waveform Encoder with Multi-Scale Globally Attentive Locally Recurrent Networks for End-to-End Speech Recognitioncited by 2
  9. 2021Exploring Heterogeneous Characteristics of Layers in ASR Models for More Efficient Trainingcited by 1
  10. 2022On the Prediction Network Architecture in RNN-T for ASRcited by 1
  11. 2023Variability of echo state network prediction horizon for partially observed dynamical systemscited by 0