Googel Gemini For Video Conference Assistance

Introduction

This project is based on my previous Kaggle competition project for Google - Gemini Long Context which require develope the AI application *Demonstrate interesting use cases for Gemini's long context window. I propose the AI Video Assistant Applciation powered by google Gemini Model capability analysis Long Context Window such as upto 1 hour long Video content.

New approach

Now I refactor this project into a fully End-to-End AI application with Localhost or Cloud service(later) : both front and backend service GUI interfaced for use implementations for Video Assistant directly without Hard coding.

As below is my original Competition link and YouTube for desciption my code for kaggle reference

Kaggle competition Links

This link for competition: https://www.kaggle.com/competitions/gemini-long-context/overview

My YouTube video Links for Kaggle

YouTube: https://youtu.be/LiKnPj5ZQjk

---

Technology use in this project

• Gemini API: It is mainlt Multimodel use for Video analysis in this project.

• FastAPI: Backend Python base framework for build Restful API servie infrasture

• React: FrontEnd Framework for building UI flow this project

• YouTube: The YouTube Video download

Installation and Setup

1. install requirements python package
2. setup Environment for Gemini API

Running project

• dev folder : for development and testing new features
• apps folder :for demonstration AI video analysis application