We use cookies. They help to improve your interaction with the site.
Electronics news

STM32F4 open source platform for tinyML projects

Maker Kingham Xu, creator of the open source OV-Watch has unveiled a new development kit: a STMicroelectronics STM32F4-based "mini devkit" for projects ranging from real-time control to tiny machine learning (tinyML): the FryPi.

The compact development kit is based on the STMicro STM32F411REU6, chosen for its compatibility with -CEU6. The board features a single 100MHz Arm Cortex-M4F microcontroller core, a digital signal processor (DSP) and ART gas pedal, 128KB of static random access memory (SRAM), and 512KB of on-chip flash memory with support for external expansion.

While the board can be used as a simple development kit by outputting its general purpose outputs (GPIOs) to two 0.1-inch dual-row pin connectors on each side, there is support for additional expansion modules, including a touchscreen display and camera sensors.

To demonstrate the microcontroller's machine learning capabilities on the device, Kingham Xu wrote a set of training programs including handwritten digit recognition, gesture recognition using a thermal camera, and face recognition using an Omnivision OV2640 sensor.

More information about the project can be found on Xu's page on Hackaday.io, and the source code and hardware design files are available on GitHub.