ST opens up the edge AI area, please visit https://stm32ai.st.com/zh/ Embedded AI is undoubtedly the next “technology hotspot”.As enterprises’ business deployment scenarios and data generation are migrating to the edge, embedded AI has also ushered in a rapid development opportunity – moving the inference process to deep edge computing brings many advantages, such as system responsiveness, user privacy protection, reduced connection costs, and power consumption.As a major driver of this trend, STMicroelectronics has invested heavily in AI to help developers quickly deploy AI applications on embedded systems based on microcontrollers/microprocessors (STM32 series) and sensors (MEMS, ToF…).
ST provides a complete set of tools to implement edge AI on STM32 MCUs, MPUs, and smart sensors, bringing intelligence to many solutions in a simple, fast, and low-cost way, such as: predictive maintenance, IoT products, smart buildings, asset tracking, people counting, etc.
Application CasesWith embedded AI, easily enhance applications, unlock new possibilities, and open the door to widespread AI applications. ST offers a wealth of application cases covering various fields such as smart cities, smart homes, entertainment, toys, smart buildings, transportation, smart offices, industry, and home appliances. Users can explore these inspiring real examples and utilize ST’s resources to create their own applications.
Products and SolutionsSTMicroelectronics offers a variety of AI solutions for STM32 and smart sensors, with various micro machine learning solutions to embed AI into microcontrollers, microprocessors, and smart sensors. Regardless of expertise in machine learning, ST’s extensive product offerings allow users to find suitable tools to meet any edge AI project needs.
NanoEdge AI Studio: Automated machine learning tool for STM32 MCUs
NanoEdge AI Studio is a user-friendly desktop tool that adds new data processing capabilities to enhance products. Any application case involving outlier/anomaly detection, classification, or using regression techniques to predict future states can harness the power of machine learning. NEAI Studio can create a custom library optimized for any STM32 in minutes for signal analysis, enhancing product intelligence.With NanoEdge AI Studio, users can easily generate machine learning libraries for embedded devices, containing millions of pre-built models. This means there is no need to collect and record large and complex datasets. Users’ models can also self-train on their devices.
STM32Cube.AI: AI model optimizer for STM32 MCUs
STM32Cube.AI is a free tool that helps optimize the performance and memory usage of trained AI models in STM32 projects. It supports TensorFlow™ Lite, Keras, and ONNX formats. If users have AI knowledge, STM32Cube.AI will automatically optimize trained artificial neural networks and generate corresponding C code for STM32 microcontrollers.STM32Cube.AI Developer Cloud PlatformThe STM32Cube.AI Developer Cloud Platform is the online version of STM32Cube.AI. It can be used to create, optimize, and generate artificial intelligence suitable for STM32 microcontrollers, as well as conduct benchmarking. No software installation or evaluation boards are required. With ST Board Farm, algorithms can even be remotely tested for performance across multiple evaluation boards.This tool is available for PC and can also be used directly online via the STM32Cube.AI Developer Cloud. This online platform provides benchmarking services to remotely evaluate AI performance on a range of STM32 boards. Additionally, access to the STM32 Model Zoo is available, which aggregates a large number of optimized AI models along with some application examples, training scripts, etc.
AI for LINUX: Complete AI framework for STM32 MPU on OpenSTLinux
For developers using STM32 MPU, X-LINUX-AI is a collection of libraries and runtime systems that simplify the integration of trained AI models in OpenSTLinux-based projects. ST has developed a complete framework for developers using OpenSTLinux to easily integrate AI models.
STM32 Hardware: Various ICs and boards for edge AI
ST provides a variety of microcontrollers, microprocessors, and smart sensors for developing edge AI applications optimized for power consumption, size, and cost.
Rich ResourcesUsers can find everything they need to understand embedded machine learning here, with links to useful content for various solutions: NanoEdge AI Studio, STM32Cube.AI, and X-LINUX-AI; find packages that integrate specific examples to easily kickstart projects.
Ready? Scan the QR code below to log in to the STM32 AI Solutions area, find the most suitable tools, upgrade your products with AI, and let’s embark on a brand new journey of AI application innovation together!
