This repository contains the code for our project developed during the HackUPC. The challenge was proposed by Inditex, where participants were given a large dataset of images and were tasked with building an algorithm to calculate similarity scores. We showcased our solution through a React frontend.
We developed a solution that utilizes llava in a RAG architecture to extract features from images and compute their similarity. The project consists of two main components: a backend developed in Python with Flask and a frontend developed in React.
- Image Similarity Calculation: Our backend algorithm calculates similarity scores between images based on extracted features.
- React Frontend: We developed a user-friendly frontend using React to showcase the results.
- Efficient Backend: The backend is built in Python using Flask, ensuring efficient processing of image data.
- Python
- Flask
- React
- chromaDB
- llava