Electronics Lab

CamThink NeoEyes NE301: Ultra-Low-Power Edge-AI Vision Camera Launched

The NeoEyes NE301 is a compact edge-AI camera with STM32N6, 0.6 TOPS NPU, Wi-Fi 6, 4MP sensor, and on-device YOLO support, designed for smart city, factory, and agriculture uses.



NeoEyes NE301 edge-AI Camera

The CamThink NeoEyes NE301 is an ultra-low-power edge-AI camera powered by the STMicro STM32N6 Arm Cortex-M55 MCU with a Neural-ART NPU offering 0.6 TOPS of compute. This enables on-device inference for lightweight tasks such as person detection, gesture recognition, and similar models without relying on cloud processing. It comes with a default 4MP MIPI-CSI sensor and also supports USB camera sensors, while providing 64MB PSRAM, 128MB HyperFlash, Wi-Fi 6, Bluetooth 5.4, audio wafers, USB-C, UART debug, and a 16-pin GPIO header. Power options include USB, battery, or PoE, with an optional 4G LTE module for US or global variants.

Designed for long-term deployment, the edge-AI camera features low power consumption, simple installation, IP67 protection, and strong expansion support. These capabilities make it suitable for Smart Cities (parking occupancy detection, smart advertising), Smart Agriculture (crop and livestock monitoring), and Smart Factory applications such as safety gear inspection and helmet detection.

Previously, we covered several other camera modules such as the EDATEC ED-AIC3100 12MP industrial smart camera with Raspberry Pi CM5, the Tokay Lite advanced edge-AI camera, and e-con Systems’ ready-to-deploy AI vision kit, among others. Feel free to check them out if you’re interested in similar products.

CamThink NeoEyes NE301 Camera

NeoEyes NE301 Camera

CamThink NeoEyes 301 specifications:

  • Core Processing:
    • MCU: STMicro STM32N6
    • CPU: Arm 32-bit Cortex-M55 @ up to 800 MHz with Arm Helium vector extensions and Arm MVE
    • GPU: Neo-Chrom 2.5D GPU with Chrom-ART Accelerator (DMA2D)
    • NPU: ST Neural-ART accelerator @ 1 GHz, up to 600 GOPS (0.6 TOPS), real-time inference, 3 TOPS/W for fanless operation
    • BPU: Hardware-accelerated H.264 and JPEG encoders up to 1080p@30fps
    • ISP: Dedicated ISP with demosaic and Auto White Balance (AWB)
  • Memory and Storage:
    • 4.2 MB SRAM
    • 64 MB PSRAM
    • 128 MB HyperFlash
    • microSD card slot
  • Camera Features:
    • MIPI CSI-2 connector with 4MP OS04C10 sensor
      • Field of View options: 51° (standard), 88° (wide), 137° (ultra-wide)
    • 4-pin USB/PIR connector for optional SC200AI camera sensor
    • Built-in fill light
  • Audio:
    • 1x Audio Input (Wafer)
    • 1x Audio Output (Wafer)
  • Wireless Connectivity:
    • Wi-Fi 6 and Bluetooth 5.4 (BLE)
    • 12-pin + 16-pin headers for optional Quectel LTE modules:
      • EG912UGL (Global, except North America)
      • EG915QNA (North America)
  • I/O and Expansion:
    • 1x USB-C port (480 Mbps)
    • 16-pin GPIO connector: UART, RS485, I2C, SPI, 2x GPIO, 5V, 3.3V, GND
    • Debug interfaces: USB-C and 4-pin UART
    • Reset, Boot, and Capture/Recording buttons
    • Power (PWR), Status (STS), and Debug LEDs
    • 2-pin wafer for alarm inputs
  • Power and Performance:
    • Power Input: 5V via USB-C
    • Battery support through 2-pin connector (4x AA supported)
    • Optional PoE via add-on
    • STM32U073Kx MCU for independent power management
    • Power Consumption:
      • Sleep: 7–8 µA (via STM32U0 power control chip)
      • Active: 170–180 mA
    • Boot / Wake: Microsecond boot, millisecond wake-up
  • Physical Specs:
    • Dimensions: 77 x 77 x 48 mm
    • Operating Temperature: –20°C to 50°C
    • Humidity: 0–95% RH (non-condensing)
    • Certifications: CE, FCC, RoHS, SRRC

CamThink NeoEyes NE301 edge-AI Camera Specifications

CamThink NeoEyes NE301 edge-AI Camera Specifications

The CamThink NeoEyes NE301 runs YOLO on-device and its firmware reveals a built-in Wi-Fi AP and a comprehensive Web user interface (UI) that allows users to preview AI inference, switch model types, modify model parameters, and fine-tune other features right in the browser. The AI vision camera also supports Supports STM32 Cube.AI, YOLOv8n, MobileNet, EfficientNet, ResNet, and more. Additional information can be found in the documentation; in particular, the “Model Training and Deployment” section provides detailed instructions for running YOLOv8 Nano Object Model along with build and flashing instructions. The open-source firmware and mechanical files (for 3D printing) are accessible.

The standard, wide, or ultra-wide angle cameras on the NeoEyes NE301 are available for pre-order for $159.90. December 2025 is also when shipping is expected to begin. Users can find additional information on the product page.

Images used courtesy of camthink.ai

Subscribe
Notify of
guest

0 Comments
Inline Feedbacks
View all comments