Milvus Image Search

A Python-based implementation of an advanced image search system utilizing Milvus vector database and image embeddings from ResNet.

Overview

This repository contains the implementation of an image search system using Milvus, a highly efficient vector database, combined with deep learning embeddings generated by a pre-trained ResNet model. The system converts images into high-dimensional vectors and uses these for fast and accurate image retrieval.

Features

• Utilizes ResNet for generating robust image embeddings.
• Employs Milvus for efficient storage and retrieval of high-dimensional image vectors.
• Supports scalable image search with advanced indexing strategies.

Prerequisites

• Python 3.8+
• PyTorch
• timm
• PIL
• pymilvus

Results