Enhancing Aviation Communication Transcription: Fine-Tuning Distil-Whisper with LoRA
Journal of Aerospace Information Systems · 2026-07 · 23 (7) · 590-596
DOI: 10.2514/1.i011656 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Aerospace Information Systems, 2026-07. Verfasst von Shokoufeh Mirzaei und 2 weiteren Autor:innen. Die Arbeit nennt 20 Literaturangaben. Gefördert von: Air Force Research Laboratory.
Abstract (Originalsprache)
Transcription of aviation communications has several applications, from assisting air traffic controllers in identifying the accuracy of read-back errors to search and rescue operations. Recent advances in artificial intelligence have provided unprecedented opportunities for improving aviation communication transcription tasks. OpenAI’s Whisper is one of the leading automatic speech recognition models. However, fine-tuning Whisper for aviation communication transcription is not computationally efficient. Thus, this paper aims to use a parameter-efficient fine-tuning method called low-rank adaptation to fine-tune a more computationally efficient version of Whisper, Distil-Whisper. To perform the fine-tuning, we used the Air Traffic Control Corpus dataset from the Linguistic Data Consortium, which contains approximately 70 h of controller and pilot transmissions near three major airports in the United States. The objective was to reduce the word error rate to enhance accuracy in the transcription of aviation communication. First, starting with an initial set of hyperparameters for Low-Rank Adaptation (LoRA) ([Formula: see text]), we performed a grid search. We applied a fivefold cross-validation to find the best combination of Distil-Whisper hyperparameters. Then, we fine-tuned the model for LoRA hyperparameters, achieving an impressive average word error rate of 3.86% across five folds. This result highlights the model’s potential for use in the cockpit.
Förderung
- Air Force Research Laboratory
Zitieren
APA
Mirzaei, S., Arzate, J. B., & Vijay, Y. (2026). Enhancing Aviation Communication Transcription: Fine-Tuning Distil-Whisper with LoRA. Journal of Aerospace Information Systems, 23(7), 590-596. https://doi.org/10.2514/1.i011656
Vancouver
Mirzaei S, Arzate JB, Vijay Y. Enhancing Aviation Communication Transcription: Fine-Tuning Distil-Whisper with LoRA. Journal of Aerospace Information Systems. 2026;23(7):590-596. doi:10.2514/1.i011656
BibTeX
@article{Mirzaei2026,
author = {Mirzaei, Shokoufeh and Arzate, Jesus B. and Vijay, Yukti},
title = {Enhancing Aviation Communication Transcription: Fine-Tuning Distil-Whisper with LoRA},
journal = {Journal of Aerospace Information Systems},
year = {2026},
volume = {23},
number = {7},
pages = {590--596},
doi = {10.2514/1.i011656}
}
EUVIMED