Multilingual Conference Audio Transcription, Speaker Diarization And Translation System

Abstract

The increasing prevalence of multilingual conferences and meetings has intensified the need for accurate communication, reliable documentation, and inclusive accessibility across diverse linguistic groups. Conventional approaches based on manual transcription, speaker identification, and translation are inefficient, costly, and susceptible to errors caused by overlapping speech, accents, background noise, and domain-specific terminology. To overcome these limitations, this work presents an AI-driven framework that integrates Automatic Speech Recognition (ASR), speaker diarization, and Neural Machine Translation (NMT). The system automatically converts spoken audio into text, identifies and segments individual speakers, and translates the content into multiple target languages while preserving semantic meaning and contextual accuracy. The resulting speaker-labeled multilingual transcripts support effective knowledge management, decision-making, and post-meeting analysis. By automating end-to-end meeting content processing, the proposed approach enhances efficiency, reduces human effort, improves accuracy, and promotes inclusive participation in academic, corporate, and international communication settings.