Stuttering-Aware Automatic Speech Recognition for Indonesian Language
This addresses the limitation of ASR systems for stuttered speech in under-represented languages like Indonesian, though it is incremental as it adapts existing methods to a new domain.
The paper tackles the problem of automatic speech recognition degrading on stuttered speech in low-resource languages like Indonesian by proposing a data augmentation framework that generates synthetic stuttered audio to fine-tune a pre-trained model, reducing recognition errors on stuttered speech while maintaining fluent performance.
Automatic speech recognition systems have achieved remarkable performance on fluent speech but continue to degrade significantly when processing stuttered speech, a limitation that is particularly acute for low-resource languages like Indonesian where specialized datasets are virtually non-existent. To overcome this scarcity, we propose a data augmentation framework that generates synthetic stuttered audio by injecting repetitions and prolongations into fluent text through a combination of rule-based transformations and large language models followed by text-to-speech synthesis. We apply this synthetic data to fine-tune a pre-trained Indonesian Whisper model using transfer learning, enabling the architecture to adapt to dysfluent acoustic patterns without requiring large-scale real-world recordings. Our experiments demonstrate that this targeted synthetic exposure consistently reduces recognition errors on stuttered speech while maintaining performance on fluent segments, validating the utility of synthetic data pipelines for developing more inclusive speech technologies in under-represented languages.