Investigation on the Robustness of Acoustic Foundation Models on Post Exercise Speech

TL;DR

This study benchmarks various acoustic foundation models for automatic speech recognition on post-exercise speech, revealing significant model-dependent robustness issues and the potential of fine-tuning to improve performance.

Contribution

It provides a comprehensive evaluation of ASR models on post-exercise speech, highlighting robustness challenges and the effectiveness of in-domain fine-tuning.

Findings

Most models degrade on post-exercise speech compared to resting speech.

FunASR achieves the strongest baseline robustness with 14.57% WER.

Fine-tuning improves CTC-based models, but Whisper shows unstable adaptation.

Abstract

Automatic speech recognition (ASR) has been extensively studied on neutral and stationary speech, yet its robustness under post-exercise physiological shift remains underexplored. Compared with resting speech, post-exercise speech often contains micro-breaths, non-semantic pauses, unstable phonation, and repetitions caused by reduced breath support, making transcription more difficult. In this work, we benchmark acoustic foundation models on post-exercise speech under a unified evaluation protocol. We compare sequence-to-sequence models (Whisper and FunASR/Paraformer) and self-supervised encoders with CTC decoding (Wav2Vec2, HuBERT, and WavLM), under both off-the-shelf inference and post-exercise in-domain fine-tuning. Across the Static/Post-All benchmark, most models degrade on post-exercise speech, while FunASR shows the strongest baseline robustness at 14.57% WER and 8.21% CER on Post-All. Fine-tuning substantially improves several CTC-based models, whereas Whisper shows unstable adaptation. As an exploratory case study, we further stratify results by fluent and non-fluent speakers; although the non-fluent subset is small, it is consistently more challenging than the fluent subset. Overall, our findings show that post-exercise ASR robustness is strongly model-dependent, that in-domain adaptation can be highly effective but not uniformly stable, and that future post-exercise ASR studies should explicitly separate fluency-related effects from exercise-induced speech variation.