Thank you for sending your enquiry! One of our team members will contact you shortly.
Thank you for sending your booking! One of our team members will contact you shortly.
Duration 21 hours
Course Outline
Introduction to TinyML
- Exploring the constraints and potential of TinyML.
- An overview of prevalent microcontroller platforms.
- A comparative analysis of Raspberry Pi, Arduino, and alternative boards.
Hardware Setup and Configuration
- Preparing the Raspberry Pi OS environment.
- Configuring Arduino boards for development.
- Interfacing with sensors and peripheral devices.
Data Collection Techniques
- Capturing data from various sensor types.
- Processing audio, motion, and environmental metrics.
- Constructing structured, labeled datasets.
Model Development for Edge Devices
- Choosing appropriate model architectures for constrained devices.
- Training TinyML models using TensorFlow Lite.
- Assessing model performance within embedded contexts.
Model Optimization and Conversion
- Applying quantization techniques.
- Converting models for microcontroller compatibility.
- Optimizing for memory usage and computational efficiency.
Deployment on Raspberry Pi
- Executing TensorFlow Lite inference on the device.
- Integrating model outputs into application logic.
- Diagnosing and resolving performance bottlenecks.
Deployment on Arduino
- Leveraging the Arduino TensorFlow Lite Micro library.
- Transferring models to microcontrollers.
- Validating accuracy and runtime behavior.
Building Complete TinyML Applications
- Architecting comprehensive embedded AI workflows.
- Implementing interactive, real-world prototypes.
- Testing and iterating on project functionality.
Conclusion and Future Directions
Requirements
- A foundational understanding of basic programming principles.
- Practical experience in utilizing microcontrollers.
- Proficiency in Python or C/C++.
Intended Audience
- Makers.
- Enthusiasts.
- Embedded AI developers.
