Pitch Accent Detection improves Pretrained Automatic Speech Recognition
Not on the map. Why?
Abstract (arXiv)
We show the performance of Automatic Speech Recognition (ASR) systems that use semi-supervised speech representations can be boosted by a complimentary pitch accent detection module, by introducing a joint ASR and pitch accent detection model. The pitch accent detection component of our model achieves a significant improvement on the state-of-the-art for the task, closing the gap in F1-score by 41%. Additionally, the ASR performance in joint training decreases WER by 28.3% on LibriSpeech, under limited resource fine-tuning. With these results, we show the importance of extending pretrained speech models to retain or re-learn important prosodic cues such as pitch accent.
Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 17 of its 44 references that are in the map. First posted on arXiv 6 Aug 2025.
Builds on
17 of 44 references.
References · 17
- 2020wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representationscited by 3,905
- 2021HuBERT: Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Unitscited by 2,342
- 2019Libri-Light: A Benchmark for ASR with Limited or No Supervisioncited by 467
- 2020Improved Noisy Student Training for Automatic Speech Recognitioncited by 141
- 2020The Zero Resource Speech Benchmark 2021: Metrics and baselines for unsupervised spoken language modelingcited by 101
- 2020Iterative Pseudo-Labeling for Speech Recognitioncited by 83
- 2020DeCoAR 2.0: Deep Contextualized Acoustic Representations with Vector Quantizationcited by 77
- 2019Effectiveness of self-supervised pre-training for speech recognitioncited by 72
- 2020SlimIPL: Language-Model-Free Iterative Pseudo-Labelingcited by 54
- 2017The Zero Resource Speech Challenge 2017cited by 45
- 2020The Zero Resource Speech Challenge 2020: Discovering discrete subword and word unitscited by 39
- 2023ProsAudit, a prosodic benchmark for self-supervised speech modelscited by 25
5 more
- 2022Detection of Prosodic Boundaries in Speech Using Wav2Vec 2.0cited by 6
- 2022A neural prosody encoder for end-ro-end dialogue act classificationcited by 5
- 2023Improving End-to-End SLU performance with Prosodic Attention and Distillationcited by 5
- 2020Data augmentation using prosody and false starts to recognize non-native children's speechcited by 3
- 2021CNN Encoding of Acoustic Parameters for Prominence Detectioncited by 1
Built on by
Citing papers per year: 2026: 1.
1 paper.