README.md

metadata

title: Traffic Sign Classification
emoji: 🐠
colorFrom: purple
colorTo: green
sdk: docker
pinned: false

Traffic Sign Classifier Flask App

This project deploys a traffic_classifier.h5 model as a Flask web app for Hugging Face Spaces with Docker.

Features

• Welcome page based on the provided visual template direction
• Login and registration
• Protected traffic sign prediction page
• SQLite storage inside the container at instance/traffic_signs.sqlite3
• Per-user prediction history
• True/false feedback for every prediction
• Dashboard with total predictions, reviewed predictions, true/false counts, and feedback accuracy

Run Locally

python -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
python app.py

Open http://localhost:7860.

Model

Place the trained model in the project root:

traffic_classifier.h5

The app expects a 43-class traffic sign classifier using 30x30 RGB images, matching the common GTSRB class list.

Hugging Face Space

This Space uses Docker and exposes port 7860.

For production, set a strong secret:

SECRET_KEY=your-secret-value

Docker Deployment

Prerequisites

• Docker installed on your system
• Docker Hub account (for pushing to registry)
• All project files including traffic_classifier.h5

Building Docker Image

Build the Docker image locally:

docker build -t traffic-sign-classifier:latest .

Running Docker Container Locally

Run the container on your local machine:

docker run -p 7860:7860 \
  -e SECRET_KEY=your-secret-key \
  -v $(pwd)/instance:/app/instance \
  traffic-sign-classifier:latest

Then access the application at http://localhost:7860.

Pushing to Docker Hub

1. Tag the image:

docker tag traffic-sign-classifier:latest yourusername/traffic-sign-classifier:latest

2. Push to Docker Hub:

docker login
docker push yourusername/traffic-sign-classifier:latest

Deploying to Hugging Face Spaces

1. Create a new Space on Hugging Face Spaces

2. Select Docker as the SDK

3. In the Space settings, set environment variable:

   • SECRET_KEY=your-production-secret

4. Upload your project files including:

   • Dockerfile
   • app.py
   • requirements.txt
   • traffic_classifier.h5
   • templates/ directory
   • static/ directory

5. Hugging Face will automatically build and deploy the container

6. Your app will be accessible at https://huggingface.co/spaces/YOUR-USERNAME/YOUR-SPACE-NAME

Docker Compose (Optional)

Create a docker-compose.yml for local development:

version: '3.8'
services:
  traffic-classifier:
    build: .
    ports:
      - "7860:7860"
    environment:
      - SECRET_KEY=dev-secret-key
      - FLASK_ENV=development
    volumes:
      - ./instance:/app/instance
      - ./templates:/app/templates
      - ./static:/app/static

Run with:

docker-compose up

Persistent Data

The SQLite database is stored in the instance/ directory, which is mounted as a volume. This ensures data persists across container restarts.

Health Check

To verify the container is running:

curl http://localhost:7860/

You should receive the welcome page HTML.

`