EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

PRL-DAS: Robust Heliox Speech Recognition for Unaligned Low-Resource Data

Yonghong Chen, Guoqi Zhang, Wanzhi Wen, Shibing Zhang

Big Data and Cognitive Computing · 15.05.2026 · 10 (5) · 157

DOI: 10.3390/bdcc10050157 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Big Data and Cognitive Computing, online erschienen am 15.05.2026. Verfasst von Yonghong Chen und 3 weiteren Autor:innen. Die Arbeit nennt 18 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Natural Science Foundation of China, Jiangsu universities, Nantong Key Laboratory of Artificial Intelligence New Quality Technology, Jiangsu College of Engineering and Technology.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Speech produced in helium–oxygen (heliox) environments in deep saturation diving exhibits pronounced spectral shifts and temporal distortions, which severely degrade automatic speech recognition (ASR) systems trained on normal-air corpora. Existing studies often adopt a restoration-then-recognition paradigm by training waveform mapping networks on paired heliox/air recordings. However, in realistic low-resource data collection, paired recordings are typically obtained by independent re-reading and are therefore not strictly time-aligned, which makes regression-style restoration more sensitive to pairing errors and increases the risk of front-end distortions. This paper proposes a robust recognition framework for heliox speech, termed PRL-DAS (Physics-informed Resampling and LoRA with Duration-Adaptive Speed). The framework consists of a physics-inspired linear resampling warm start (PhysSpeed), parameter-efficient Low-Rank Adaptation (LoRA), and duration-adaptive speed (DAS) inference enhancement. Specifically, we first apply physics-motivated linear resampling as a coarse warm start, and then perform mixed-domain LoRA fine-tuning of a Whisper foundation model to absorb residual non-linear differences. On a corpus of 1048 paired Chinese heliox utterances under leave-one-speaker-out (LOSO) evaluation, using Whisper-Medium as the base model, PhysSpeed followed by mixed-domain LoRA reduces the overall character error rate (CER) from 49.33% with PhysSpeed preprocessing only to 25.79%, while also improving performance on the normal domain. Furthermore, the full PRL-DAS framework applies Soft-DAS, a lightweight smooth schedule motivated by duration-dependent variation in the optimal resampling factor, and further reduces the overall CER to 24.37% without additional training cost.

Förderung

  • National Natural Science Foundation of China
  • Jiangsu universities
  • Nantong Key Laboratory of Artificial Intelligence New Quality Technology, Jiangsu College of Engineering and Technology

Zitieren

APA

Chen, Y., Zhang, G., Wen, W., & Zhang, S. (2026). PRL-DAS: Robust Heliox Speech Recognition for Unaligned Low-Resource Data. Big Data and Cognitive Computing, 10(5), 157. https://doi.org/10.3390/bdcc10050157

Vancouver

Chen Y, Zhang G, Wen W, Zhang S. PRL-DAS: Robust Heliox Speech Recognition for Unaligned Low-Resource Data. Big Data and Cognitive Computing. 2026;10(5):157. doi:10.3390/bdcc10050157

BibTeX
@article{Chen2026,
  author = {Chen, Yonghong and Zhang, Guoqi and Wen, Wanzhi and Zhang, Shibing},
  title = {PRL-DAS: Robust Heliox Speech Recognition for Unaligned Low-Resource Data},
  journal = {Big Data and Cognitive Computing},
  year = {2026},
  volume = {10},
  number = {5},
  pages = {157},
  doi = {10.3390/bdcc10050157}
}
Kostenlos starten →