# ManaTTS Persian: a recipe for creating TTS datasets for lower resource languages

arXiv: https://arxiv.org/abs/2409.07259 · HTML: https://mlsanity.com/paper/2409.07259/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 11 Sep 2024.

## Built on by (most cited first, 8 of 8)

- [LLM-Powered Grapheme-to-Phoneme Conversion: Benchmark and Case Study](https://mlsanity.com/paper/2409.08554/) (2024)
- [ParsVoice: A Large-Scale Multi-Speaker Persian Speech Corpus for Text-to-Speech Synthesis](https://mlsanity.com/paper/2510.10774/) (2025)
- [Fast, Not Fancy: Rethinking G2P with Rich Data and Rule-Based Models](https://mlsanity.com/paper/2505.12973/) (2025)
- [Efficient ASR for Low-Resource Languages: Leveraging Cross-Lingual Unlabeled Data](https://mlsanity.com/paper/2512.07277/) (2025)
- [PARSA-Bench: A Comprehensive Persian Audio-Language Model Benchmark](https://mlsanity.com/paper/2603.14456/) (2026)
- [Beyond Unified Models: A Service-Oriented Approach to Low Latency, Context Aware Phonemization for Real Time TTS](https://mlsanity.com/paper/2512.08006/) (2025)
- [Correlation between prosody and pragmatics: A case study of the discourse marker h=al=a `now' in Persian](https://mlsanity.com/paper/2607.28359/) (2026)
- [Neyshekar: An Open Persian Read-Speech Corpus for Automatic Speech Recognition](https://mlsanity.com/paper/2609.14542/) (2026)

## Abstract (arXiv)

In this study, we introduce ManaTTS, the most extensive publicly accessible single-speaker Persian corpus, and a comprehensive framework for collecting transcribed speech datasets for the Persian language. ManaTTS, released under the open CC-0 license, comprises approximately 86 hours of audio with a sampling rate of 44.1 kHz. Alongside ManaTTS, we also generated the VirgoolInformal dataset to evaluate Persian speech recognition models used for forced alignment, extending over 5 hours of audio. The datasets are supported by a fully transparent, MIT-licensed pipeline, a testament to innovation in the field. It includes unique tools for sentence tokenization, bounded audio segmentation, and a novel forced alignment method. This alignment technique is specifically designed for low-resource languages, addressing a crucial need in the field. With this dataset, we trained a Tacotron2-based TTS model, achieving a Mean Opinion Score (MOS) of 3.76, which is remarkably close to the MOS of 3.86 for the utterances generated by the same vocoder and natural spectrogram, and the MOS of 4.01 for the natural waveform, demonstrating the exceptional quality and effectiveness of the corpus.