ST, STM32Cube.AI version 7.0 released for free
Adding traditional machine learning algorithms on top of neural networks
Processing of various AI models is possible on STM32 MCUs


STMicroelectronics announced on the 29th that it has expanded the machine learning techniques available in the STM32Cube.AI development environment. Now, we can efficiently solve tasks such as classification, clustering, and novelty detection with more flexibility.

STM32Cube.AI version 7.0 enables the development of neural networks for edge inference on STM32 MCUs, supporting new supervised and semi-supervised approaches that operate on smaller data sets and with fewer CPU cycles.

It includes the K-means and SVM classifier algorithms for classification, isolation forest (iForest) for outlier detection, and OC SVM (One Class Support Vector Machine) that can be implemented without complex manual coding.

Traditional machine learning algorithms are added on top of neural networks, enabling developers to quickly convert, validate, and deploy various types of AI models on STM32 MCUs.

STM32Cube.AI users can run machine learning workloads in the cloud on STM32-based edge devices, minimizing data exchange over the Internet. This reduces latency, saves energy, improves cloud utilization, and protects privacy.

STM32Cube.AI version 7.0 is now available for free download from www.st.com .