Guardado en:
| Autor principal: | |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | https://arxiv.org/abs/2512.00937 |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866914175015452672 |
|---|---|
| author | Nippert, Lars |
| author_facet | Nippert, Lars |
| contents | Arabic text-to-speech (TTS) remains challenging due to limited resources and complex phonological patterns. We present reproducible baselines for Arabic TTS built on the FastPitch architecture and introduce cepstral-domain metrics for analyzing oversmoothing in mel-spectrogram prediction. While traditional Lp reconstruction losses yield smooth but over-averaged outputs, the proposed metrics reveal their temporal and spectral effects throughout training. To address this, we incorporate a lightweight adversarial spectrogram loss, which trains stably and substantially reduces oversmoothing. We further explore multi-speaker Arabic TTS by augmenting FastPitch with synthetic voices generated using XTTSv2, resulting in improved prosodic diversity without loss of stability. The code, pretrained models, and training recipes are publicly available at: https://github.com/nipponjo/tts-arabic-pytorch. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_00937 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Arabic TTS with FastPitch: Reproducible Baselines, Adversarial Training, and Oversmoothing Analysis Nippert, Lars Audio and Speech Processing Arabic text-to-speech (TTS) remains challenging due to limited resources and complex phonological patterns. We present reproducible baselines for Arabic TTS built on the FastPitch architecture and introduce cepstral-domain metrics for analyzing oversmoothing in mel-spectrogram prediction. While traditional Lp reconstruction losses yield smooth but over-averaged outputs, the proposed metrics reveal their temporal and spectral effects throughout training. To address this, we incorporate a lightweight adversarial spectrogram loss, which trains stably and substantially reduces oversmoothing. We further explore multi-speaker Arabic TTS by augmenting FastPitch with synthetic voices generated using XTTSv2, resulting in improved prosodic diversity without loss of stability. The code, pretrained models, and training recipes are publicly available at: https://github.com/nipponjo/tts-arabic-pytorch. |
| title | Arabic TTS with FastPitch: Reproducible Baselines, Adversarial Training, and Oversmoothing Analysis |
| topic | Audio and Speech Processing |
| url | https://arxiv.org/abs/2512.00937 |