🖼️ Image Classification with Teachable Machine

This project demonstrates how to train a real-time image classification model without writing any code, using Google's Teachable Machine. It classifies images into two categories using webcam or uploaded inputs and displays prediction confidence instantly.

---

🔍 Objective

To create a no-code machine learning model that can accurately distinguish between two visual classes (e.g., apples vs bananas, thumbs up vs thumbs down) using Teachable Machine’s browser-based platform.

---

📁 Dataset

• Source: Self-collected via webcam and/or uploaded images
• Classes: Example — Apple and Banana
• Size: ~30 images per class (recommended for training)

You can also use curated image datasets from sources like Unsplash, Pexels, or Kaggle Fruits 360.

---

🛠️ Tools & Platforms

• Teachable Machine by Google
• TensorFlow.js (optional for export)

---

💡 Key Learnings

• Teachable Machine enables rapid prototyping for image classification tasks
• No coding required — great for beginners, students, and educators
• Models can be exported and embedded in real-world applications

---

📁 Folder Structure

├── Apples/ # Class 1 images
├── Bananas/ # Class 2 images
├── converted_savedmodel/ # Exported TensorFlow model
├── LICENSE
└── README.md