StutterCut: Uncertainty-Guided Normalised Cut for Dysfluency Segmentation
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866916879476457472 |
|---|---|
| author | Ghosh, Suhita Jouaiti, Melanie Perschewski, Jan-Ole Stober, Sebastian |
| author_facet | Ghosh, Suhita Jouaiti, Melanie Perschewski, Jan-Ole Stober, Sebastian |
| contents | Detecting and segmenting dysfluencies is crucial for effective speech therapy and real-time feedback. However, most methods only classify dysfluencies at the utterance level. We introduce StutterCut, a semi-supervised framework that formulates dysfluency segmentation as a graph partitioning problem, where speech embeddings from overlapping windows are represented as graph nodes. We refine the connections between nodes using a pseudo-oracle classifier trained on weak (utterance-level) labels, with its influence controlled by an uncertainty measure from Monte Carlo dropout. Additionally, we extend the weakly labelled FluencyBank dataset by incorporating frame-level dysfluency boundaries for four dysfluency types. This provides a more realistic benchmark compared to synthetic datasets. Experiments on real and synthetic datasets show that StutterCut outperforms existing methods, achieving higher F1 scores and more precise stuttering onset detection. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2508_02255 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | StutterCut: Uncertainty-Guided Normalised Cut for Dysfluency Segmentation Ghosh, Suhita Jouaiti, Melanie Perschewski, Jan-Ole Stober, Sebastian Sound Artificial Intelligence Audio and Speech Processing Detecting and segmenting dysfluencies is crucial for effective speech therapy and real-time feedback. However, most methods only classify dysfluencies at the utterance level. We introduce StutterCut, a semi-supervised framework that formulates dysfluency segmentation as a graph partitioning problem, where speech embeddings from overlapping windows are represented as graph nodes. We refine the connections between nodes using a pseudo-oracle classifier trained on weak (utterance-level) labels, with its influence controlled by an uncertainty measure from Monte Carlo dropout. Additionally, we extend the weakly labelled FluencyBank dataset by incorporating frame-level dysfluency boundaries for four dysfluency types. This provides a more realistic benchmark compared to synthetic datasets. Experiments on real and synthetic datasets show that StutterCut outperforms existing methods, achieving higher F1 scores and more precise stuttering onset detection. |
| title | StutterCut: Uncertainty-Guided Normalised Cut for Dysfluency Segmentation |
| topic | Sound Artificial Intelligence Audio and Speech Processing |
| url | https://arxiv.org/abs/2508.02255 |