ACES: Accent Subspaces for Coupling, Explanations, and Stress-Testing in Automatic Speech Recognition
Fuente:
arXiv
Salvato in:
| Autore principale: | |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866917320964702208 |
|---|---|
| author | Parekh, Swapnil |
| author_facet | Parekh, Swapnil |
| contents | ASR systems exhibit persistent performance disparities across accents, but whether these gaps reflect superficial biases or deep structural vulnerabilities remains unclear. We introduce ACES, a three-stage audit that extracts accent-discriminative subspaces from ASR representations, constrains adversarial attacks to them, and tests whether removing them improves fairness. On Wav2Vec2-base with seven accents, imperceptible perturbations (~60 dB SNR) along the accent subspace amplify the WER disparity gap by nearly 50% (21.3->31.8 pp), exceeding random-subspace controls; a permuted-label test confirms specificity to genuine accent structure. Partially removing the subspace worsens both WER and disparity, revealing that accent-discriminative and recognition-critical features are deeply entangled. ACES thus positions accent subspaces as powerful fairness-auditing tools, not simple erasure levers. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_03359 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | ACES: Accent Subspaces for Coupling, Explanations, and Stress-Testing in Automatic Speech Recognition Parekh, Swapnil Sound Artificial Intelligence Audio and Speech Processing ASR systems exhibit persistent performance disparities across accents, but whether these gaps reflect superficial biases or deep structural vulnerabilities remains unclear. We introduce ACES, a three-stage audit that extracts accent-discriminative subspaces from ASR representations, constrains adversarial attacks to them, and tests whether removing them improves fairness. On Wav2Vec2-base with seven accents, imperceptible perturbations (~60 dB SNR) along the accent subspace amplify the WER disparity gap by nearly 50% (21.3->31.8 pp), exceeding random-subspace controls; a permuted-label test confirms specificity to genuine accent structure. Partially removing the subspace worsens both WER and disparity, revealing that accent-discriminative and recognition-critical features are deeply entangled. ACES thus positions accent subspaces as powerful fairness-auditing tools, not simple erasure levers. |
| title | ACES: Accent Subspaces for Coupling, Explanations, and Stress-Testing in Automatic Speech Recognition |
| topic | Sound Artificial Intelligence Audio and Speech Processing |
| url | https://arxiv.org/abs/2603.03359 |