FluencyBank

FluencyBank is a dataset that contains recordings of fluent and disfluent speech used to evaluate the performance of speech processing models, particularly in the context of stuttering.

Papers using FluencyBank (3)

Paediatric-HGNN: A Hybrid Heterogeneous Graph Neural Network for Detecting Disfluency in Children's Speech via Multiscale Acoustic Fusion2026StutterZero and StutterFormer: End-to-End Speech Conversion for Stuttering Transcription and Correction2025Optimizing Multi-Stuttered Speech Classification: Leveraging Whisper's Encoder for Efficient Parameter Reduction in Automated Assessment2024