# Arabic TTS with FastPitch: Reproducible Baselines, Adversarial Training, and Oversmoothing Analysis

arXiv: https://arxiv.org/abs/2512.00937 · HTML: https://mlsanity.com/paper/2512.00937/

## Key facts

- Cited by 1 papers in the map (data through 19 Oct 2026).
- Placed in Multilingual and low-resource ASR, one of 11 directions in Speech and audio language models.
- Builds on 13 of its 28 references that are in the map.
- First posted on arXiv 30 Nov 2025.

## Placed in

- Speech and audio language models / [Multilingual and low-resource ASR](https://mlsanity.com/speech/multilingual-and-low-resource-asr/)

## Builds on (most cited first, 13 of 13)

- [Spectral Normalization for Generative Adversarial Networks](https://mlsanity.com/paper/1802.05957/) (2018)
- [HiFi-GAN: Generative Adversarial Networks for Efficient and High Fidelity Speech Synthesis](https://mlsanity.com/paper/2010.05646/) (2020)
- [Least Squares Generative Adversarial Networks](https://mlsanity.com/paper/1611.04076/) (2016)
- [Natural TTS Synthesis by Conditioning WaveNet on Mel Spectrogram Predictions](https://mlsanity.com/paper/1712.05884/) (2017)
- [Conditional Variational Autoencoder with Adversarial Learning for End-to-End Text-to-Speech](https://mlsanity.com/paper/2106.06103/) (2021)
- [Free-Form Image Inpainting with Gated Convolution](https://mlsanity.com/paper/1806.03589/) (2018)
- [XTTS: a Massively Multilingual Zero-Shot Text-to-Speech Model](https://mlsanity.com/paper/2406.04904/) (2024)
- [Vocos: Closing the gap between time-domain and Fourier-based neural vocoders for high-quality audio synthesis](https://mlsanity.com/paper/2306.00814/) (2023)
- [FastPitch: Parallel Text-to-speech with Pitch Prediction](https://mlsanity.com/paper/2006.06873/) (2020)
- [One TTS Alignment To Rule Them All](https://mlsanity.com/paper/2108.10447/) (2021)
- [Fast Spectrogram Inversion using Multi-head Convolutional Neural Networks](https://mlsanity.com/paper/1808.06719/) (2018)
- [Neural Arabic Text Diacritization: State of the Art Results and a Novel Approach for Machine Translation](https://mlsanity.com/paper/1911.03531/) (2019)
- [CATT: Character-based Arabic Tashkeel Transformer](https://mlsanity.com/paper/2407.03236/) (2024)

## Built on by (most cited first, 1 of 1)

- [An Ultra-Low Latency, End-to-End Streaming Speech Synthesis Architecture via Block-Wise Generation and Depth-Wise Codec Decoding](https://mlsanity.com/paper/2604.12438/) (2026)

## Abstract (arXiv)

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.