Our teams just published a Massive Open Online Course that serves as an Introduction to STM32Cube.AI, the industry’s most advanced AI toolkit capable of converting neural networks into optimized code for STM32 MCUs. Until now, engineers could use our demo applications and boards as a stepping stone toward their own systems. With these courses, they can now get a deeper understanding of the subject matter itself and how it will apply to their particular situations. The coursework still uses our Function Pack FP-AI-SENSING1 to introduce the various fundamental concepts that developers need to master, such as labeling data or inference operations. The video below shows that even enthusiasts with no prior experience in AI can use the demo applications to get a working understanding of the basic blocks that come together to create a machine-learning system.
Introduction to STM32Cube.AI: Neural Network Creation
Students will also appreciate the fact that the MOOC delves into the creation of a neural network. STM32Cube.AI assumes that developers already have a neural network and are looking to use it on an MCU to run inferences. However, the course goes into the creation of such networks using Keras and Python. The lecturer already has a set of data, and he shows how to train models, what can influence accuracies, or how to use a confusion matrix to evaluate the behavior of a model. It obviously won’t replace the work of a data scientist, but it is crucial information to developers who will better understand what they can expect from their data scientist and what they should look for.
Introduction to STM32Cube.AI: Function Calls and Weights Representations
Finally, the MOOC offers a lab that will assist students as they port their models onto an STM32 code. They will understand how STM32CubeMX can hasten developments and how the tool can take a model and convert it. They will also get a behind-the-scenes look at the
MX_X_CUBE_AI_Process function that the end application will call and how it runs the model as a meta program thanks to an interesting representation of the different weights in an array. This is a particularly important concept because it helps developers understand how STM32Cube.AI greatly simplifies the implementation of machine learning mechanisms on embedded systems.