Dysarthria Normalization via Local Lie Group Transformations for Robust ASR
Fuente:
arXiv
Saved in:
| Main Author: | |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909600641449984 |
|---|---|
| author | Osipov, Mikhail |
| author_facet | Osipov, Mikhail |
| contents | We present a geometry-driven method for normalizing dysarthric speech by modeling time, frequency, and amplitude distortions as smooth, local Lie group transformations of spectrograms. Scalar fields generate these deformations via exponential maps, and a neural network is trained - using only synthetically warped healthy speech - to infer the fields and apply an approximate inverse at test time. We introduce a spontaneous-symmetry-breaking (SSB) potential that encourages the model to discover non-trivial field configurations. On real pathological speech, the system delivers consistent gains: up to 17 percentage-point WER reduction on challenging TORGO utterances and a 16 percent drop in WER variance, with no degradation on clean CommonVoice data. Character and phoneme error rates improve in parallel, confirming linguistic relevance. Our results demonstrate that geometrically structured warping provides consistent, zero-shot robustness gains for dysarthric ASR. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_12279 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Dysarthria Normalization via Local Lie Group Transformations for Robust ASR Osipov, Mikhail Sound Computation and Language Machine Learning Audio and Speech Processing We present a geometry-driven method for normalizing dysarthric speech by modeling time, frequency, and amplitude distortions as smooth, local Lie group transformations of spectrograms. Scalar fields generate these deformations via exponential maps, and a neural network is trained - using only synthetically warped healthy speech - to infer the fields and apply an approximate inverse at test time. We introduce a spontaneous-symmetry-breaking (SSB) potential that encourages the model to discover non-trivial field configurations. On real pathological speech, the system delivers consistent gains: up to 17 percentage-point WER reduction on challenging TORGO utterances and a 16 percent drop in WER variance, with no degradation on clean CommonVoice data. Character and phoneme error rates improve in parallel, confirming linguistic relevance. Our results demonstrate that geometrically structured warping provides consistent, zero-shot robustness gains for dysarthric ASR. |
| title | Dysarthria Normalization via Local Lie Group Transformations for Robust ASR |
| topic | Sound Computation and Language Machine Learning Audio and Speech Processing |
| url | https://arxiv.org/abs/2504.12279 |