Cyklops: Video/Image to Text Conversational Agent

Cyklops is an intelligent conversational agent that provides video/image-to-text capabilities. It utilizes the FastAPI framework along with the open-source LLaVA model for efficient image-to-caption processing. This repo contains the frontend of Cyklops based on Next JS . For backend , visit Cyklops-Backend

Demo Video

demo.mp4

Getting Started

First , install the dependencies using

npm install

Second , make sure to add LLaVA API and Backend Endpoint in the api file

const LLAVA_API_URL = " "; //Your ngrok or hosted LLaVA model url here
const FASTAPI_API_URL = ""; //Your FastAPI Url Here
const github_repo = ``; //Your username/repo which you will work as storage bucket

Now, run the development server:

npm run dev
# or
yarn dev
# or
pnpm dev
# or
bun dev

Open http://localhost:3000 with your browser to see the result.

You are all set , Visualise , Mesmerise and Converse with Videos like never before

Proudly open source