bfredd9 / meta-st-stm32mpu-ai

This repository contains the OpenEmbedded meta layer to install AI frameworks and tools for the STM32MP1

meta-st-stm32mpu-ai

OpenEmbedded meta layer to install AI frameworks and tools for the STM32MP1. It also provide application samples.

Compatibility

This version has been validated against the OpenSTLinux ecosystem release v2.1.0, v2.0.0 and v1.2.0. It supports STM32MP157x-DKx, STM32MP157x-EV1 and Avenger96 boards.

Available frameworks and tools within the meta-layer

X-LINUX-AI v2.1.0 expansion package:

• TensorFlow™ Lite 2.4.1
• armNN 20.11
• OpenCV 4.1.x
• Python™ 3.8.x (enabling Pillow module)
• Support STM32MP15xF devices operating at up to 800MHz
• Support of the Avenger96 board from Linaro™ 96Boards based on the STM32MP157A microprocessor, either with a USB camera or the DesignCore® OV5640 camera mezzanine board from D3 Engineering tested with the OpenSTLinux Distribution v2.1.0
• Coral Edge TPU accelerator support
  • libedgetpu 2.4.1 (built from source) aligned with TensorFlow™ Lite 2.4.1
• The X-LINUX-AI OpenSTLinux Expansion Package v2.1.0 is compatible with Yocto Project® build systems Thud and Dunfell. As a consequence, it is compatible with OpenSTLinux Distributions v1.2.0, v2.0.0 and v2.1.0 on STM32MP157C-DK2 with a USB camera, and on STM32MP157A-EV1 and STM32MP157C-EV1 with their built-in camera module
• Support for the OpenSTLinux AI package repository allowing the installation of prebuilt package using apt-* utilities
• Application samples
  • C++ / Python™ image classification application using TensorFlow™ Lite based on MobileNet v1 quantized model
  • C++ / Python™ object detection application using TensorFlow™ Lite based on COCO SSD MobileNet v1 quantized model
  • C++ / Python™ image classification application using Coral Edge TPU™ based on MobileNet v1 quantized model and compiled for the Coral Edge TPU
  • C++ / Python™ object detection applicationusing Coral Edge TPU™ based on COCO SSD MobileNet v1 quantized model and compiled for the Coral Edge TPU
  • C++ image classification application using armNN TensorFlow™ Lite parser based on MobileNet v1 float model
  • C++ object detection application using armNN TensorFlow™ Lite parser based on COCO SSD MobileNet v1 quantized model
  • C++ face recognition application using proprietary model capable of recognizing the face of a known (enrolled) user. Contact the local STMicroelectronics support for more information about this application or send a request to [e-mail]

Further information on how to install and how to use

https://wiki.st.com/stm32mpu/wiki/X-LINUX-AI_OpenSTLinux_Expansion_Package

Application samples

https://wiki.st.com/stm32mpu/wiki/X-LINUX-AI_application_samples_zoo