EMNLP Findings2024
GE2PE: Persian End-to-End Grapheme-to-Phoneme Conversion
Abstract
Text-to-Speech (TTS) systems have made significant strides, enabling the generation of speech from grapheme sequences. However, for low-resource languages, these models still struggle to produce natural and intelligible speech. Grapheme-to-Phoneme conversion (G2P) addresses this challenge by enhancing the input sequence with phonetic information. Despite these advancements, existing G2P systems face limitations when dealing with Persian texts due to the complexity of Persian transcription. In this study, we focus on enriching resources for the Persian language. To achieve this, we introduce two novel G2P training datasets: one manually labeled and the other machine-generated. These datasets comprise over five million sentences alongside their corresponding phoneme sequences. Additionally, we propose two evaluation datasets tailored for Persian sub-tasks, including Kasre-Ezafe detection, homograph disambiguation, and handling out-of-vocabulary (OOV) words. To tackle the unique challenges of the Persian language, we develop a new sentence-level Endto-End (E2E) model leveraging a two-step training approach, as outlined in our paper, to maximize the impact of manually labeled data. The results show that our model surpasses the stateof-the-art performance by 1.86% in word error rate, 4.03% in Kasre-Ezafe detection recall, and 3.42% in homograph disambiguation accuracy. 1
Ask Lune
Know this paper in minutes, not an afternoon.
Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.
Lineage
Builds on2
- Zero-shot Learning for Grapheme to Phoneme Conversion with Language EnsembleXinjian Li, Florian Metze, David R. Mortensen, Shinji Watanabe et al.ACL Findings 2022 · 12 citations
- Persian Ezafe Recognition Using Transformers and Its Role in Part-Of-Speech TaggingEhsan Doostmohammadi, Minoo Nassajian, Adel RahimiEMNLP Findings 2020 · 6 citations
Related papers
- Fast, Not Fancy: Rethinking G2P with Rich Data and Statistical ModelsMahta Fetrat Qharabagh, Zahra Dehghanian, Hamid R. RabieeEMNLP Findings 2025 · 1 citation
- Spanning the Spectrum of Hatred Detection: A Persian Multi-Label Hate Speech Dataset with Annotator RationalesZahra Delbari, Nafise Sadat Moosavi, Mohammad Taher PilehvarAAAI 2024 · 11 citations
- GigaSpeech 2: An Evolving, Large-Scale and Multi-domain ASR Corpus for Low-Resource Languages with Automated Crawling, Transcription and RefinementYifan Yang, Zheshu Song, Jianheng Zhuo, Mingyu Cui et al.ACL 2025
- Deepfake Defense: Constructing and Evaluating a Specialized Urdu Deepfake Audio DatasetSheza Munir, Wassay Sajjad, Mukeet Raza, Emaan Abbas et al.ACL Findings 2024 · 7 citations
- Dict-TTS: Learning to Pronounce with Prior Dictionary Knowledge for Text-to-SpeechZiyue Jiang, Su Zhe, Zhou Zhao, Qian Yang et al.NeurIPS 2022 · 7 citations
