Typical vs. Atypical Disfluency Classification: Introducing the IIITH-TISA Corpus and Temporal Context-Based Feature Representations

TL;DR

This paper introduces the IIITH-TISA corpus for atypical disfluencies, proposes a novel feature set with a TDNN classifier, and achieves high accuracy in classifying typical versus atypical speech disfluencies.

Contribution

The study presents the first Indian English stammer corpus and a new feature extraction method for improved disfluency classification.

Findings

Achieved an average F1 score of 85.01% in disfluency classification.

Introduced the IIITH-TISA dataset for atypical disfluencies in Indian English.

Demonstrated the effectiveness of PE-ZTWCC and SDC features with TDNN.

Abstract

Speech disfluencies in spontaneous communication can be categorized as either typical or atypical. Typical disfluencies, such as hesitations and repetitions, are natural occurrences in everyday speech, while atypical disfluencies are indicative of pathological disorders like stuttering. Distinguishing between these categories is crucial for improving voice assistants (VAs) for Persons Who Stutter (PWS), who often face premature cutoffs due to misidentification of speech termination. Accurate classification also aids in detecting stuttering early in children, preventing misdiagnosis as language development disfluency. This research introduces the IIITH-TISA dataset, the first Indian English stammer corpus, capturing atypical disfluencies. Additionally, we extend the IIITH-IED dataset with detailed annotations for typical disfluencies. We propose Perceptually Enhanced Zero-Time Windowed Cepstral Coefficients (PE-ZTWCC) combined with Shifted Delta Cepstra (SDC) as input features to a shallow Time Delay Neural Network (TDNN) classifier, capturing both local and wider temporal contexts. Our method achieves an average F1 score of 85.01% for disfluency classification, outperforming traditional features.