Nuvoton NuMaker-GestureAI-M55M1 module combines Cortex-M55 MCU with GC0308 camera for AI gesture control
A $30 module from Nuvoton bundles a Cortex-M55 MCU, Ethos-U55 NPU, and a VGA camera to recognize over ten hand gestures on the device itself.
Condensed by AI-Portable from Editorial queue.
A Tight Stack of AI-Ready Silicon
Nuvoton’s NuMaker-GestureAI-M55M1 module is built around the NuMicro M55M1R2LJAE MCU, which pairs an Arm Cortex-M55 running at 220 MHz with the Ethos-U55 microNPU, also clocked at 220 MHz. On‑chip memory includes 1.5 MB of SRAM and 2 MB of flash, giving plenty of headroom for local inference without external RAM. A microSD card slot on the back stores AI model files, while a GC0308 VGA CMOS image sensor captures 640×480 frames and an on‑board digital microphone handles audio. Connectivity is straightforward: a USB‑C port for power and high‑speed data, plus expansion headers for debugging and UART/I2C output. The board is rated for -40°C to +105°C, so it’s comfortable in industrial environments.
Over Ten Gestures from the Factory
Out of the box, the module ships with firmware that tracks human presence and classifies hand gestures. The pre‑trained set includes:
- Call
- First
- Like
- Mute
- OK
- One
- Palm
- Peace
- Stop
- Three
Detection results stream over the UART5 interface in a simple serial protocol: [AA][ID][x y w h][55][CC]. The ID byte encodes a person (0x00–0x7F) or a gesture (0x80–0x8A), while four decimal fields give bounding box coordinates from 0 to 9999. This makes integration with host microcontrollers or single‑board computers trivial—just parse the hex.
Custom Models with TensorFlow Lite
Nuvoton backs the hardware with an AI toolchain that supports TensorFlow Lite for Microcontrollers. Developers can swap in their own models for gesture recognition, object detection, or even audio tasks using the microphone. The firmware and source code are available on GitHub, and the product page walks through getting started. At $30, the NuMaker-GestureAI-M55M1 sits in the sweet spot for prototyping touchless interfaces, smart appliances, or portable vision systems—though at the time of writing, it’s temporarily out of stock on Nuvoton’s direct store.
The module represents a growing trend: bringing edge AI capabilities into physically tiny, low‑cost packages without relying on cloud connectivity. With the Cortex‑M55 and Ethos‑U55 tandem, it delivers respectable on‑device inference for a class of applications where privacy, latency, or lack of network access rules out remote processing.
Back in November last year, we covered the launch of the NuMicro M55M1 MCU from Nuvoton , which combines an Arm Cortex-M55 core with an Ethos-U55 microNPU for on-device AI and gesture control. Now, they have released the NuMaker-GestureAI-M55M1, a development module based on that MCU for AI vision-related applications.
This new board integrates the M55M1 MCU with a GC0308 CMOS image sensor, a digital microphone, and a microSD card slot for storing AI models. It is designed for applications such as gesture control, basic vision systems, and touchless interfaces.
MCU – Nuvoton NuMicro M55M1R2LJAE CPU – Cortex-M55 MCU @ 220 MHz Memory – 1.5 MB SRAM Storage – 2 MB Flash AI Accelerator – Arm Ethos-U55 micro-NPU @ 220 MHz