FoodVision AI (Food Image Classification)

Project Overview

An end-to-end deep learning computer vision system that classifies images into 11 distinct food categories using transfer learning and deep convolutional neural networks.

Key Features

• Transfer Learning Architecture: Fine-tuned a pre-trained VGG16 convolutional neural network on curated food image benchmarks to achieve high classification accuracy.
• RESTful API Backend: Developed a high-performance backend using FastAPI to process uploaded images and return prediction confidence scores.
• Interactive UI: Built a responsive React interface allowing real-time image uploads, instant inference preview, and confidence score visualization.
• Containerization: Packaged the complete application stack using Docker for consistent, cross-platform deployment.

Tech Stack

• Deep Learning: Python, VGG16, Scikit-learn
• Backend & Web: FastAPI, React, Docker