Image Classification API using FastAPI and ResNet50

This is a FastAPI project for building an image classification API using the ResNet50 model. The API allows users to upload an image and receive predictions about its content.

Setup and Installation

1. Install Dependencies: Ensure you have Python installed on your machine. Install the required Python packages using the following: pip install fastapi uvicorn[standard] pillow numpy tensorflow

2. Run the API: Save the provided code in a file (e.g., app.py). Open a terminal and run the following command: uvicorn app:app --reload The API will be accessible at http://127.0.0.1:8000.

Usage

1. Home Endpoint: Navigate to http://127.0.0.1:8000/ in your web browser or use a tool like curl to test: curl http://127.0.0.1:8000/

2. Prediction Endpoint: Use the /predict endpoint to upload an image and receive predictions. You can use curl as follows: curl -X POST http://127.0.0.1:8000/predict -H "Content-Type: multipart/form-data" -F "file=@path/to/your/image.jpg"

Example Usage (requests library)

import requests

# Specify the API endpoint
api_url = "http://127.0.0.1:8000/predict"

# Open the image file
files = {'file': ('lion.jpg', open('path/to/your/lion.jpg', 'rb'))}

# Make the prediction request
response = requests.post(api_url, files=files)

# Print the response
print(response.json())

API USAGE (with React)