Video Captioning Studio

Video Captioning Studio is an AI-powered web application that transforms uploaded videos into engaging, captioned content. Users can upload an MP4 video, and the system automatically analyzes both the audio and on-screen visuals using Google's Gemini multimodal AI. Based on the video's content, the application generates captions and summaries in the distinct style chosen by the user (Formal, Sarcastic, Humorous-Tech, Humorous-Non-Tech, and a personalized option). A key feature of the platform is its caption personalization system. Users can create their own custom caption style by writing a short prompt (up to 50 characters) and selecting one or more existing caption styles as inspiration. The AI then combines the selected styles with the user's prompt to generate a unique, personalized captioning style tailored to their preferences. For example, formal + sarcastic = dry wit. The platform uses FFmpeg to create captioned videos, with each style displayed using a distinct caption color for easy identification. Generated captions, summaries, processing history, and output files are stored in a SQLite database, allowing users to access and review previous results in the history page easily. Built with a FastAPI backend and a React frontend. The project makes video content more accessible, informative, entertaining, and personalized by enabling users to experiment with different caption tones and create entirely new styles of their own.