Have you ever wanted to try facial and/or voice recognition? The ESP-EYE d'Espressif is a miniature development card based on the ESP32 that combines a digital microphone, the ESP32 (of course) with 8 MB of PSRAM memory and 4 MB of flash, and a 2 megapixel camera. There are also some buttons and LEDs for basic control and configuration.

Unlike some cards, this code database does not require an Internet connection - you do not need to send video or audio data to the cloud - everything is processed on the chip! The integrated demo shows what he can do:

    Plug it into a computer's USB port and open a 115,200 baud serial port connection to the new serial port.
    Once the red LED lights up, activate the card with the password "Hi Lehshin" (in the documents they spell it lexin, the x is pronounced as a shh sound)
    The red LED flashes and then connects to the new Wi-Fi access point created.
    Open a web browser at http://192.168.4.1/face_stream
    Point the camera at the faces to show the recognition boxes! The video is about 2 frames per second, not super fast.

Once the demo is launched, check the source code and ESP development platform to customize your own voice and DIY speech recognition project.

Specifications :

    WiSoC ESP32 dual core processor Tensilica LX6 with WiFi and Bluetooth
    Memory: 8 MB PSRAM
    Storage: 4MB Flash
    Camera: 2MP OV2640 Camera
    Audio: MEMS Microphone
    USB: 1 microUSB port for power supply and serial port programming
    Various Reset, start and function buttons, 2x LEDs
    Dimensions: 41 x 21 mm