mlsanityData to 7 Oct 2026

The RWTH ASR System for TED-LIUM Release 2: Improving Hybrid HMM with SpecAugment

arXiv 2004.00960 ↗First posted 2 Apr 2020eess.AS cs.SDcited by 19

Not on the map. Why?

Abstract (arXiv)

We present a complete training pipeline to build a state-of-the-art hybrid HMM-based ASR system on the 2nd release of the TED-LIUM corpus. Data augmentation using SpecAugment is successfully applied to improve performance on top of our best SAT model using i-vectors. By investigating the effect of different maskings, we achieve improvements from SpecAugment on hybrid HMM models without increasing model size and training time. A subsequent sMBR training is applied to fine-tune the final acoustic model, and both LSTM and Transformer language models are trained and evaluated. Our best system achieves a 5.6% WER on the test set, which outperforms the previous state-of-the-art by 27% relative.

Key facts. Cited by 19 papers in the map (data through 7 Oct 2026). First posted on arXiv 2 Apr 2020.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 8; 2022: 3; 2023: 3; 2024: 2; 2025: 1; 2026: 2.

19 papers.

Sort:

  1. 2021Robust wav2vec 2.0: Analyzing Domain Shift in Self-Supervised Pre-Trainingcited by 148
  2. 2023End-to-End Speech Recognition: A Surveycited by 127
  3. 2021SpeechStew: Simply Mix All Available Speech Recognition Data to Train One Large Neural Networkcited by 83
  4. 2021Transformer Based Deliberation for Two-Pass Speech Recognitioncited by 29
  5. 2022Listen, Adapt, Better WER: Source-free Single-utterance Test-time Adaptation for Automatic Speech Recognitioncited by 19
  6. 2021Conformer-based Hybrid ASR System for Switchboard Datasetcited by 15
  7. 2022Efficient Training of Neural Transducer for Speech Recognitioncited by 14
  8. 2023Lexical Speaker Error Correction: Leveraging Language Models for Speaker Diarization Error Correctioncited by 14
  9. 2021Acoustic Data-Driven Subword Modeling for End-to-End Speech Recognitioncited by 11
  10. 2022Improving the Training Recipe for a Robust Conformer-based Hybrid Modelcited by 9
  11. 2021On Architectures and Training for Raw Waveform Feature Extraction in ASRcited by 8
  12. 2024Dynamic Encoder Size Based on Data-Driven Layer-wise Pruning for Speech Recognitioncited by 7
7 more
  1. 2024AG-LSEC: Audio Grounded Lexical Speaker Error Correctioncited by 3
  2. 2021The Impact of ASR on the Automatic Analysis of Linguistic Complexity and Sophistication in Spontaneous L2 Speechcited by 1
  3. 2021Prediction of Listener Perception of Argumentative Speech in a Crowdsourced Dataset Using (Psycho-)Linguistic and Fluency Featurescited by 1
  4. 2025Dynamic Acoustic Model Architecture Optimization in Training for ASRcited by 1
  5. 2023Hybrid ASR for Resource-Constrained Robots: HMM - Deep Learning Fusioncited by 0
  6. 2026Do we really need Self-Attention for Streaming Automatic Speech Recognition?cited by 0
  7. 2026SENS-ASR: Semantic Embedding injection in Neural-transducer for Streaming Automatic Speech Recognitioncited by 0