Console Controller Classifier

A browser-based web app interface for classifying images of game console controllers. Upload an image and the app sends it to a backend ML model trained to recognize 4 different console controller types. This lightweight front-end showcases the model's classification capabilities with real-time predictions.

What it is

This project is a frontend interface that:

• accepts image file uploads (JPG, PNG, etc.)
• sends images to a backend ML model API for classification
• displays classification results with confidence scores for each of the 4 console controller types

Features

• lightweight static web app (no dependencies)
• drag-and-drop image upload with preview
• real-time classification results with confidence percentages
• connected to backend ML model via API
• responsive, modern UI

Usage

1. Open index.html in a browser.
2. Upload an image of a console controller (drag-and-drop or click to browse).
3. Click the "Identify" button to classify the image.
4. View the classification results with confidence scores for each of the 4 controller types.

Backend API

The app requires a backend ML model API endpoint. Update the API constant in index.html to point to your backend or use mine:

const API = 'https://medo-id-testing.hf.space/classify'

The backend should accept POST requests with JSON payload { data: "<base64_image>" } and return classification results as JSON with controller types and confidence scores.

Files

• index.html — main web page and app UI

Model Training

This app is designed to showcase a trained ML model that recognizes:

1. PS Controller
2. Xbox Controller
3. Nintendo Controller
4. Gamecube Controller

Good for

• showcasing trained ML models
• demos of computer vision classification
• interactive model testing interface
• prototyping browser-based classifiers