Lingo-Sync: Real-Time Multilingual Translation System for Seamless Global Communication Using AI and M2M-100
Year: 2025
Type: article
Abstract: LingoSync is a multilingual AI-driven translation platform designed to bridge global communicationgaps by providing accurate, real-time translation across over 100 language pairs. Unlike traditional systemsthat use English as an intermediate language, LingoSync leverages Facebook’s M2M100 model to performdirect language-to-language translation, significantly enhancing translation quality and reducing contextualdistortion. The system features a lightweight, web-based interface built using Flask, supporting real-timetranslation, automatic source language detection, and a conversation mode for live multilingualcommunication.In addition to text-based translation, LingoSync integrates Optical Character Recognition (OCR) toextract translatable content from images and scanned documents, expanding its utility to non-textual media.Its modular architecture supports translation across multiple formats, including speech, images, anddocuments, making it suitable for diverse use cases such as education, administration, travel, and crossborder collaboration. Rigorous evaluation has confirmed the system’s effectiveness in both translationaccuracy and computational efficiency, making LingoSync a scalable and practical solution for real-worldmultilingual communication.
Author: Shubham Pedhe, Shreya Kudale, Pratham Mali, Aarsh Sharma, Prof. Nutan Dholzake
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W6893420193
FWCI:
Cites:
Cited by:
Related to: