Edge AI
Reading list · Learning path
| Textbook | Volume I: Introduction to Machine Learning Systems |
| Prerequisites | Python and introductory machine learning |
| Format | No semester plan yet |
| Hardware | A Raspberry Pi kit |
| Materials | Slides · Edge AI Engineering, a companion book by Marcelo Rovai |
Description
Students deploy models on Linux edge devices that sit close to the sensors and people that use the results, so memory, power, and response time limit the design. The path pairs Volume I chapters with experiments on a Raspberry Pi.
Reading list
| Reading | Slides |
|---|---|
| ML Lifecycle | Ch 3 |
| Model Compression | Ch 10 |
| Hardware Acceleration | Ch 11 |
| Model Serving | Ch 13 |
Hands-on experiments
| Raspberry Pi kit lab |
|---|
| Setup |
| Image Classification |
| Object Detection |
| Small Language Models |
| Vision-Language Models |
The companion book’s weekly lab sequence offers one possible order. Each experiment needs a test run on the classroom hardware before it is assigned.
Hardware
Approximate prices from the Raspberry Pi kit page; a camera module is extra. The language and vision-language model experiments need a Raspberry Pi 5.
| Board | Approximate price |
|---|---|
| Raspberry Pi Zero 2 W | ~$15 |
| Raspberry Pi 4 (4 GB) | ~$55 |
| Raspberry Pi 5 (4 GB) | ~$60 |
| Raspberry Pi 5 (8 GB) | ~$80 |
Status
The path has no semester syllabus yet. The Edge AI course page lists courses at other institutions.