Ending the Blind Flight: Analyzing the Impact of Acoustic and Lexical Factors on WAV2VEC 2.0 in Air-Traffic Control.

Transformer neural networks have shown remarkable success on standard automatic speech recognition (ASR) benchmarks. However, they are known to be less robust against domain mismatch, particularly with air traffic control (ATC) speech data. In the ATC domain, transformer-based ASR systems do usually not transfer across different datasets. The reasons for poor transferability across ATC datasets remain unclear. Our study investigates the influence of acoustic variability and lexical differences on the ASR performance across various ATC datasets. By fine-tuning and evaluating wav2vec 2.0 on synthetic ATC datasets, we examine the effect of acoustic variability on the model performance. Furthermore, we assess the effect of lexical differences by correlating language model perplexity with performance. Our findings reveal that a combination of acoustic and lexical mismatch causes the bad inter-dataset transferability and give insights on how to improve future ASR models for ATC.