Gravity: Offline Edge AI Gesture & Face Detection Sensor – 5 Gestures, 10 Faces, 3m

SKU: SEN0626

The Gravity: Offline Edge AI Gesture & Face Detection Sensor is a versatile, privacy-focused module designed for intuitive human-computer interaction, featuring an onboard AI processor that performs all recognition locally—without any cloud dependency—to ensure real-time responsiveness and complete data security. This powerful sensor offers dual core functionalities: it accurately recognizes five common hand gestures (including thumbs-up and open palm) with instant colored LED feedback, and simultaneously detects the presence of up to 10 people by identifying their head-and-shoulders area, making it ideal for touchless control systems, occupancy analytics, and interactive installations. With an 85° field of view and a 3-meter range, it delivers reliable performance in varied environments, while its wide 3.3V–5V voltage support and dual I2C/UART interfaces ensure full compatibility with Arduino, Raspberry Pi, ESP32, and graphical programming platforms, all within a low-power design perfect for standalone and smart IoT projects.

Overview of Gesture and Face Detection Module
  • Docs
  • Tech Specs
  • 1. Example Code for Arduino-Reading Data via I2C

    Reading Data via I2C

    2. Example Code for Arduino-Reading Data via UART

    Reading Data via UART

    3. Example Code for Arduino-Dazzling LED Ring

    Create a dazzling, gesture-controlled LED ring with the Gravity AI Sensor. The ring lights up when a person is detected and changes color based on five distinct hand gestures (e.g., "Good" for Blue, "OK" for Green), offering an intuitive, touchless interactive lighting experience.

    4. Example Code for Arduino-Face-Tracking Fan

    This project creates a smart, face-tracking fan using the Gravity AI Sensor. It turns on/off with hand gestures ("OK" to start, "Open Palm" to stop) and automatically follows your face's horizontal position in real-time, keeping you cool without any physical contact.

    5. Example Code for Mind+-Graphical Programming

    No coding is required. By dragging and dropping graphical building blocks, you can make the sensor output the number of faces, position coordinates, and the types of recognized gestures (such as "thumbs-up", "OK", etc.).

    6. Example Code for micro:bit-MakeCode Programming

    On the micro:bit development board, program with graphical building blocks to enable the sensor to detect the presence of a face, recognize gesture types, and display the results on the device, which is suitable for fun interactive projects with micro:bit.

    7. Modbus Register Definitions

    Modbus Register Definitions for Gravity: Offline Edge AI Gesture & Face Detection Sensor

    Reference

    Specification

    Dimension

    Parameter Details
    Power Parameters Supply Voltage: 3.3V-5V
    Level Voltage: 3.3V
    Operating Current: 100mA
    Interface Parameters Communication Interface: PH2.0-4P/2.54mm pin header holes
    Communication Methods: I2C/UART I2C/UART
    Default Address: 0x72
    Default UART Baud Rate: 9600
    UART Communication Protocol: modbus
    Interrupt Interface: 2.54mm pin header hole
    Recognition Parameters Types of Gesture Recognition: 5
    - Good ?
    - OK ?
    - Stop ?
    - Victory ✌️
    - Call me ?
    Recognition Distance (Gesture/Face): 0.5-3 meters
    Facial Recognition Range: Area above the human shoulders
    Maximum Number of Recognizable Faces: Up to 10
    Facial Recognition Score: 0-100
    Facial Position Coordinate Range: 0-640
    Interrupt Output: Low level output upon gesture recognition
    Camera Field of View: D=85°
    Camera Focal Length: 1.56mm
    Display Parameters Gesture Indicator Lights: RGB lights
    - Blue for ?
    - Green for ?
    - Red for ?
    - Yellow for ✌️
    - Purple for ?
    Facial Indicator Light: Single LED (lit = presence detected, unlit = no presence)

    Pinout

    Label Name Function Description
    D/T I2C data line SDA or UART transmission terminal TXD (3.3V level)
    C/R I2C clock line SCL or UART reception terminal RXD (3.3V level)
    GND Power supply negative pole
    VCC Power supply positive pole (3.3~5.5V)
    INT Interrupt output, low level output upon recognition of any gesture
    I2C/UART I2C/UART communication mode selection switch
    Head-Shoulder Indicator Light Illuminates steadily when the camera detects a face, extinguishes when the person leaves
    Gesture Indicator Light Illuminates in colors corresponding to the 5 predefined gestures

    Was this article helpful?

    TOP