← Back to the map
Beyond the course.
The tutorials take a model beyond the course, one for each intermediate course. Each one has a demo to build, the hardware you need, the complete code and how long it takes.
-
From the lab to your own page
- The demo
- The page uses the camera to recognize the classes of your Photo model, with no server and no account.
- Hardware
- A phone, or a computer with a webcam.
- Code
- HTML and JavaScript, with TensorFlow.js and MediaPipe.
- Time
- 40 minutes
-
A camera at the bird feeder
- The demo
- A Raspberry Pi with a camera saves a photo when a bird lands on the feeder, and writes down which bird it is.
- Hardware
- Raspberry Pi 5 with Camera Module 3, a power supply and a microSD card.
- Code
- Python, with MediaPipe and LiteRT.
- Time
- 2 hours
-
A sensor that hears when the washing machine is done
- The demo
- A small board listens for the end-of-cycle beep and turns on an LED.
- Hardware
- Arduino Nano 33 BLE Sense Rev2, which has a built-in microphone.
- Code
- C++ with LiteRT for Microcontrollers, or Edge Impulse.
- Time
- 2 hours
-
The magic wand, gestures on a microcontroller
- The demo
- You draw a letter in the air with the board in your hand, and it recognizes it.
- Hardware
- Arduino Nano 33 BLE Sense Rev2, which has an accelerometer and a gyroscope.
- Code
- C++ with LiteRT for Microcontrollers.
- Time
- 90 minutes
-
An assistant that answers from your notes
- The demo
- A page searches your notes by meaning and a small language model answers, all inside the browser.
- Hardware
- A computer or a recent phone with WebGPU.
- Code
- JavaScript with Transformers.js.
- Time
- 60 minutes