Vismo

What it does

Our team has developed a Smart YouTube Video Assistant Application (Vismo) that allows users to input a YouTube video URL along with a custom prompt or query. Based on the intent of the prompt, the system, leveraging the Gemini API, provides a tailored response. This response could be in the form of plain text, text with relevant images, video snippets from the specified video, or even a video response.

The application begins by extracting the video’s captions and title. With the help of the Gemini API, it generates text responses, classifies images and video snippets, or creates a script for a video summary. The app uses timestamps and captions from the transcript to accurately identify and extract relevant images and video segments. Additionally, the application enhances the user experience by offering recommendations such as web sources, related images, and YouTube videos. The Gemini API plays a crucial role in generating the search queries that fuel these web results and recommendations.