International Conference on Computer and Knowledge Engineering

Home / 14th International Conference on Computer and Knowledge Engineering

AvashoG2P: A multi-module G2P Converter for Persian

Authors :

Ali Moghadaszadeh¹ Fatemeh Pasban² Mohsen Mahmoudzadeh³ Maryam Vatanparast⁴ Amirmohammad Salehoof⁵

1- Part AI Research Center 2- Part AI Research Center 3- Ferdowsi University of Mashhad 4- Part AI Research Center 5- Part AI Research Center

Keywords :

TTS،G2P

Abstract :

The conversion of graphemes to phonemes (G2P) is a fundamental task in text-to-speech (TTS) and automatic speech recognition (ASR) systems. Over the years, G2P systems have evolved from rule-based and statistical methods to advanced neural network-based approaches. Despite these advancements, G2P conversion for Persian remains challenging due to the complex relationship between spelling and pronunciation and the scarcity of high-quality datasets. This paper introduces the AvashoG2P, a multi-module novel solution for Persian G2P conversion. The AvashoG2P system leverages a sequence-to-sequence (seq2seq) model with a GRU-based recurrent unit and an attention mechanism. This model is trained on both diacritized and non-diacritized words, enhancing its understanding of phonemes and their relationships. The system achieves a Word Error Rate (WER) of 15\% and a Phoneme Error Rate (PER) of 5\%, demonstrating its effectiveness. One of the critical components of AvashoG2P is its homograph disambiguation module, which utilizes a single model for all homographs, addressing a significant challenge in Persian text processing. Our method leverages a classification approach for homograph disambiguation, which assigns a phoneme label to the entire input window. Our system achieves high accuracy while optimizing for latency and memory consumption. We achieve significant improvements in accuracy and F1 scores using transformer-based models and machine learning classifiers. Our results highlight the superior performance of the XLMRoberta model among transformer models, with an F1 Weighted score of 94.7, and the SVC model among machine learning classifiers, with an F1 Weighted score of 89.96. Additionally, we present the AvashoG2P-Benchmark, a comprehensive test dataset designed to facilitate future research and benchmarking in Persian G2P tasks (available at: https://huggingface.co/datasets/PartAI/AvashoG2P-Benchmark).

Papers List

List of archived papers

Improvement of Credit Scoring by LSTM Autoencoder Model

Milad Sattari Maleki - Seyedeh Niusha Motevallian - Faezehsadat Hosseini - Mohammad Sabokrou - Hamidreza Soltanalizadeh Maleki

A Vision-Based Method for Human Activity Recognition Using Local Binary Pattern

Babak Goodarzi - Reza Javidan - Mohammad Sadegh Rezaei

Assessing Users' Influence on Respondents in Conversation Quality: A Quantitative Study on Reddit Based on the Cooperative Principle

Afsaneh Habibi - Fattaneh Taghiyareh

Brain Age Estimation with Twin Vision Transformer using Hippocampus Information Applicable to Alzheimer Dementia Diagnosis

Zahra Qodrati - Seyedeh Masoumeh Taji - Amirhossein Ghaemi - Habibollah Danyali - Kamran Kazemi - Alireza Ghaemi

Designing an IT2 Fuzzy Rule-based System for Emotion Recognition Using Biological Data

Mahsa Keshtkar - Hooman Tahayori

Evaluating the Impact of Traveling on COVID-19 Prevalence and Predicting the New Confirmed Cases According to the Travel Rate Using Machine Learning: A Case Study in Iran

Anita Ghandehari - Soheil Shirvani - Hadi Moradi

Deep Inside Tor: Exploring Website Fingerprinting Attacks on Tor Traffic in Realistic Settings

Amirhossein Khajehpour - Farid Zandi - Navid Malekghaini - Mahdi Hemmatyar - Naeimeh Omidvar - Mahdi Jafari Siavoshani

Automated software design using Machine Learning With Natural Language Processing

Fahimeh Khedmatkon - Seyed Mohammad Hossein Hasheminejad - Jaleh Shoshtarian Malak

FarCQA: A Farsi Community Dataset for Question Classification and Answer Selection

Saba Emami - Maedeh Mosharraf

Optimizing Question-Answering Framework Through Integration of Text Summarization Model and Third-Generation Generative Pre-Trained Transformer

Ervin Gubin Moung - Toh Sin Tong - Maisarah Mohd Sufian - Valentino Liaw - Ali Farzamnia - Farashazillah Yahya