Learning TinyML: A Hands-On Course
1h 11mIntermediate2022-07-21
Authors

Vaidheeswaran Archana
Data Scientist, AI Engineer, and Data Product Manager
Course details
While you may not realize it, TinyML probably affects your life in some way on a daily basis. If you have a smartphone or IoT device that features voice activation, facial recognition, audio detection, or other functions that employ machine learning algorithms, you have TinyML to thank. In this course, instructor Vaidheeswaran Archana guides you into the world of TinyML and shows you how you can process huge AI models right in the palm of your hand. Vaidheeswaran starts by teaching you how to identify if your ML/AI problem is a TinyML problem, then shows you optimization techniques to fit your deep learning models and illustrates multiple use cases. She explains quantization techniques, how to train a model using Tflite, how to deploy a TinyML model, and covers the entire TinyMLOps lifecycle. Vaidheeswaran finishes the course with a look at what’s in store for the future of TinyML, along with some resources you can use to continue your learning.
Skills covered
TensorFlowMachine LearningGoogleArtificial Intelligence (AI)Learning
Concepts
0. Introduction
- 01 - Getting started with TinyML
- 02 - What is TinyML
- 03 - What you should know
1. Is Your Problem a TinyML Problem
- 04 - Defining constraints
- 05 - Checklist for a TinyML problem
2. Solving the Constraints - Optimization Techniques
- 06 - Pre-trained models
- 07 - Quantization and types of quantization
- 08 - TFLite post training quantization
- 09 - Quantization awareness training in TFLite
- 10 - Pruning
- 11 - Knowledge distillation
- 12 - Challenge - Compare results of optimization
- 13 - Solution - Compare results of optimization
3. Deploying TinyML Models
- 14 - Edge impulse
- 15 - Deploy a classification project to your phone
- 16 - Challenge - Deploy a regression model to your phone
- 17 - Solution - Deploy a regression model to your phone
4. TinyMLOps
- 18 - Hardware devices
- 19 - Bringing all the concepts together
Conclusion
- 20 - Resources for TinyML
- 21 - Future of the TinyML - Research directions
- 22 - Next steps with TinyML