Overview
Explore the world of tinyML applications in this comprehensive tutorial from the tinyML Summit 2021. Learn how to build industrial-grade embedded machine learning solutions using Edge Impulse. Follow along as experts guide you through the entire process, from data collection and model design to deployment on microcontrollers. Discover best practices for creating accurate and efficient tinyML models, including dataset creation, model training, performance evaluation, and optimization for embedded systems. Gain hands-on experience with the Thunderboard kit or observe the process to understand the intricacies of integrating machine learning into real-world embedded applications. Cover topics such as sensor data processing, validation, quantization, classification, anomaly detection, and model drift. By the end of this 1 hour and 37 minute session, acquire the skills to develop and deploy your own tinyML applications for various industrial use cases.
Syllabus
Introduction
Questions
Project Setup
Adding Sensor Data
Processing Data
Validation Results
Quantizing Results
Gazelle Classification
Anomaly Detection
Model Drift
Deploy to device
Impulse source code
Demo
Conclusions
Taught by
tinyML