ARTIFICIAL INTELLIGENCE
PROJECT: "Build Your Own AI-Powered App!"
Theme: AI in My Pocket – Train, Test, and Build Your Own AI App!
Objective: To understand how machine learning models are created using Teachable Machine, and how they
can be integrated into a simple mobile app using MIT App Inventor.
Project Steps:
Step 1: Choose a Classification Project
Pick any one of the following fun projects to create using Google Teachable Machine:
● Recognize 3 different hand gestures
● Identify 3 emotions on your face (happy, sad, surprised)
● Classify 3 sounds (clap, whistle, snap)
● Detect 3 objects (like bottle, pen, phone)
Step 2: Train a Model Using Teachable Machine
● Visit: https://teachablemachine.withgoogle.com
● Choose: Image Project or Audio Project
● Add three classes and upload/record examples for each (at least 20 per class).
● Click on Train Model
● After training, click Export Model → Choose TensorFlow.js → Click "Upload (Host online)"
● Copy the model URL
Step 3: Build a Mobile App with MIT App Inventor + Teachable Machine
● Visit: https://ai2.appinventor.mit.edu
● Start a new project
● Import the Teachable Machine Extension:
○ Download from here: https://github.com/mit-cml/extensions/releases (or your teacher may
provide the .aix file)
○ Go to Extension → Import extension → Upload the .aix file
● Add Components:
○ Camera or ImagePicker to capture input
○ Teachable Machine extension
○ Label to show prediction
○ Button to trigger prediction
○ Web Viewer (optional, for previewing the model)
● Configure the Extension:
○ Set Model URL to the URL from Teachable Machine
○ Use the Call Teachable Machine. Predict Image block
○ Display results in a label
● Test your App using the MIT AI2 Companion on your phone
Step 4: Document Your Work
Use Adobe Express to design a webpage explaining the project with screenshots:
○ What is your app’s purpose?
○ How did you train your model?
○ How accurate was your model during testing?
○ What challenges did you face?
○ How can this app be useful in real life?
IBM SKILLSBUILD LEARNING ACTIVITY:
https://students.yourlearning.ibm.com/activity/PLAN-A49AC6D42035
Make Short notes in your Notebook
Class IX Revision
1. AI Domain
2. AI Project Cycle
3. Application of NLP
4. Application of Data Science
5. Application of Computer Vision
6. Supervised vs Unsupervised vs Reinforce Learning
7. Classification vs Clustering
8. AI Bias
9. AI Privacy
Project Work
(project file)
1. How can we create a script bot for our school? Provide a detailed plan and the steps required to build a script bot
for the school. Also, create a sample script bot for the school.
