Aim

To train a computer to recognize two objects or gestures using Google’s Teachable Machine.

Materials Needed

• A computer or laptop with internet access

• A webcam (built-in or external)

• Access to the website: https://teachablemachine.withgoogle.com/

• Notebook for recording observations

Step-by-Step Instructions

Step 1: Open the Website

Go to Teachable Machine and click “Get Started.”

Choose “Image Project” → “Standard Image Model.”

Step 2: Create Classes

You’ll see boxes labeled “Class 1” and “Class 2.”
Rename them — for example:

• Class 1: “Thumbs Up”

• Class 2: “Thumbs Down”
  (You can also choose objects like “Book” vs. “Pencil.”)

Step 3: Capture Images

Use your webcam to take 20–30 photos for each class.

Try to vary the lighting, background, and angle — this helps AI learn better.
Example:

• Thumbs Up in bright light

• Thumbs Up near window

• Thumbs Down in dim light

Step 4: Train Your Model

Click “Train Model.”
The computer will analyze your images and find patterns.
It may take a few seconds.

Step 5: Test the AI

Click “Preview” to test your model.
Show your thumbs up or thumbs down to the webcam.
Watch how the AI identifies your gesture and gives a percentage (confidence level).

Step 6: Observe and Record

Write down:

• Which gestures were recognized correctly?

• When did it get confused?

• How could you improve accuracy (e.g., add more training images, better lighting)?