Enhancing Multilingual Video Accessibility for Diverse Audiences

Abstract

With the rise of the internet and the evolution of digital communication, video content has emerged as a dominant communication medium. The widespread use of video content has led to multiple accessibility challenges, including those faced by individuals with hearing impairments and those with language barriers [1]. This project addresses these challenges by developing an automated software solution to enhance the accessibility of video content through translation, transcription, subtitling, and dubbing. The system employs a microservices architecture to ensure modularity, scalability, and maintainability. Each microservice handles specific tasks such as video transcription, subtitle generation, and translation. In the implementation phase, key functionalities were developed and rigorously tested. The subtitling feature allows users to upload videos of various lengths and sizes, which are then transcribed and subtitled. Transcribely supports over 120 languages, enabling users to access video content in their preferred language. Additionally, a chatbot feature powered by large language models allows users to interact with the video content, providing text summaries and answering specific questions based on the video's content. Comprehensive testing, including unit tests, component tests, and user tests, was conducted to verify the reliability and robustness of the system. Despite its capabilities, Transcribely has limitations. It relies entirely on internet connectivity, which limits accessibility in regions with poor internet infrastructure. Furthermore, while the system supports a wide range of languages, the quality of translation and transcription can vary depending on the quality of the video's audio.