Text to Video of Various PIB Press Releases Using Artificial Intelligence in English and Regional Languages

Abstract

The project "Text to Video of Various PIB Press Releases Using Artificial Intelligence in English and Regional Languages" is an AI-based system that automatically converts Press Information Bureau (PIB) press releases into informative videos in both English and regional languages by integrating text, visuals, voice narration, and subtitles to make government information more engaging and accessible. Existing PIB press releases are primarily text-based, making them difficult for many users to read and understand, while manual video creation is time-consuming, costly, and requires technical expertise, resulting in delays in information dissemination. To overcome these challenges, the proposed system automates the text-to-video generation process using Artificial Intelligence by summarizing press releases, translating the content into regional languages, generating relevant visuals, producing natural voice narration, and creating synchronized videos with subtitles. The project is developed using Python and employs Natural Language Processing , Deep Learning, Text-to-Speech, Neural Machine Translation, and Generative Artificial Intelligence. It utilizes BART for text summarization, MarianMT for multilingual translation, CLIP and BLIP for image-text understanding, and Coqui TTS for speech synthesis. The proposed system enables faster dissemination of multilingual government information, improves accessibility, reduces manual effort, and enhances public awareness through accurate and engaging AI-generated videos.