Sumit Kumar
Published © MIT

Recognizing MNIST-based Handwritten Digits on M5Stack Core2

Learn how to quickly develop a TinyML model to recognize drawn digits on touch interfaces with low-power MCUs.

BeginnerFull instructions provided3 hours250

Things used in this project

Hardware components

M5Stack Core2 ESP32 IoT Development Kit
M5Stack Core2 ESP32 IoT Development Kit
The MCU is an ESP32 model D0WDQ6-V3 and has dual-core Xtensa® 32-bit 240Mhz LX6 processors that can be controlled separately. Wi-Fi is supported as standard and it includes an onboard 16MB Flash and 8MB PSRAM, USB TYPE-C interface for charging, downloading of programs and serial communication, a 2.0-inch integrated capacitive touch screen, and a built-in vibration motor.

Software apps and online services

Neuton Tiny ML Neuton
Automatically build extremely tiny and explainable models without any coding and machine learning background and embed them into any microcontroller
Arduino IDE
Arduino IDE


Read more




Source Code


Sumit Kumar

Sumit Kumar

27 projects • 67 followers
18 y/o. My daily routine involves dealing with electronics, code, distributed storage and cloud APIs.