Has anyone integrated ai models in to micropython?

 Integrating AI models into MicroPython can be challenging due to the limited resources available on microcontrollers, which usually have constraints on memory, processing power, and energy consumption. However, there are examples of AI models being integrated into MicroPython, particularly for TinyML applications.

One approach is to use TensorFlow Lite for Microcontrollers, a version of TensorFlow Lite designed specifically for microcontrollers and embedded devices. TensorFlow Lite for Microcontrollers supports a subset of operations available in the full TensorFlow Lite, making it suitable for running on resource-constrained devices.

While TensorFlow Lite for Microcontrollers does not have direct support for MicroPython, you can use the C++ API to integrate the AI models into your MicroPython application. Some microcontrollers, like the ESP32, support both MicroPython and the TensorFlow Lite for Microcontrollers C++ API. In such cases, you can use the C++ API to run the AI model and then interact with it through MicroPython.

An example of this approach can be found in the following GitHub repository: https://github.com/mocleiri/tensorflow-micropython-examples

It demonstrates how to use TensorFlow Lite for Microcontrollers with MicroPython on an ESP32 microcontroller. The repository includes examples for running speech recognition and image classification models.

Keep in mind that running AI models on microcontrollers can be challenging due to the limited resources available. You may need to optimize the model, use quantization, or apply other techniques to make the AI model suitable for deployment on such devices.

How to setup and use ESP32 WiFi Camera

ampy - Utility to interact with a MicroPython board over a serial connection.

 

This can be really useful to control the I/O hardware if your control board is running on MicroPython like the Micro:Bit. 

I found ampy was being used by the Hadabot Project used for teaching ROS2. that consisted of an ESP32 Wifi board running ampy, and a simple DC motor controller on a magic chassis , turtle bot robot that is just two motors, a motor driver , and encoders, and a few sensors.  Everything was being over Wifi to be controlled by ROS via the ampy python library which is a cinch to in then map ROS messages in to micropython function calls.  

https://github.com/scientifichackers/ampy

ampy

MicroPython Tool (ampy) - Utility to interact with a CircuitPython or MicroPython board over a serial connection.

Ampy is meant to be a simple command line tool to manipulate files and run code on a CircuitPython or MicroPython board over its serial connection. With ampy you can send files from your computer to the board's file system, download files from a board to your computer, and even send a Python script to a board to be executed.

Note that ampy by design is meant to be simple and does not support advanced interaction like a shell or terminal to send input to a board. Check out other MicroPython tools like rshell or mpfshell for more advanced interaction with boards.

$5 Wi-Fi IP Surveillance Camera (Very little DIY needed)

https://github.com/espressif/arduino-esp32

https://www.diymore.cc/collections/esp8266/products/esp32-cam-wifi-wireless-module-esp32-serial-to-wifi-esp32-cam-spi-flash-bluetooth-development-board-with-ov2640-camera-module

https://www.aliexpress.com/item/32983165020.html

ESP32-CAM-MB CH340G ESP8266 ESP32-S Wireless WiFi Bluetooth Expansion Board OV2640 2MP Camera SMD IPEX Antenna for Arduino MCU

[ESP32-CAM-MB CH340G Serial Micro USB]

Description:

The ESP32-CAM-MB module is a small camera module with a size of 39.8*27*. This module can work independently as the smallest system. A new WiFi+Bluetooth dual-mode development board based on ESP32 design, using PCB on-board antenna, with 2 high-performance 32-bit LX6CPU, using 7-level pipeline architecture, main frequency adjustment range 80MHz to 240Mhz. Ultra-low power consumption, deep sleep current is as low as 6mA. It is an ultra-small 802.11b/g/n Wi-Fi + BT/BLE SoC module

ESP32-CAM-MB adopts Micro USB interface, convenient and reliable connection method, which is convenient to be applied to various IoT hardware terminal occasions.

 

Parameters:

Using low-power dual-core 32-bit CPU, it can be used as an application processor.

The main frequency is up to 240MHz, and the computing power is up to 600 DMIPS

Built-in 520 KB SRAM, external 8MB PSRAM

Support UART/SPI/I2C/PWM/ADC/DAC and other interfaces

Support OV2640 and OV7670 cameras, built-in flash

Support picture WiFI upload

Support TF card

Support multiple sleep modes.

Embedded Lwip and FreeRTOS

Support STA/AP/STA+AP working mode

Support Smart Config/AirKiss one-click network configuration

Support secondary development

 

Application:

Home smart device image transmission

Wireless monitoring

Smart agriculture

QR wireless recognition

Wireless positioning system signal

And other IoT applications

 

Specification:

Working voltage: 4.75 -5.25V

SPIFlash: 32Mbit

RAM: Internal 520KB + external 8MB PSRAM

Wi-Fi: 802.11b/g/n/e/i

Bluetooth: Bluetooth 4.2BR/EDR and BLE standard

Support interface: (2Mbps) UART, SPI, I2C, PWM

Support TF card: up to 4G

IO port: 9

Serial port rate: 115200bps

Spectrum range: 2400 -2483.5MHz

Antenna form: onboard PCB antenna, gain 2dBi

Image output format: JPEG (only OV2640 support), BMP, GRAYSCALE

Transmit power:

802.11b: 17±2dBm (@11Mbps)

802.11g: 14±2dBm (@54Mbps)

802.11n: 13±2dBm (@MCS7)

Receiving sensitivity:

CCK, 1Mbps: -90dBm

CCK, 11Mbps: -85dBm

6Mbps(1/2BPSK): -88dBm

54Mbps(3/464-QAM): -70dBm

MCS7 (65Mbps, 72.2Mbps): -67dBm

Power consumption:

Flash off: 180mA@5V

Turn on the flash and adjust the brightness to the maximum: 310mA@5V

Deep-sleep: The lowest power consumption can reach 6mA@5V

Moderm-sleep: The minimum can reach 20mA@5V

Light-sleep: The lowest can reach 6.7mA@5V

Security: WPA/WPA2/WPA2-Enterprise/WPS

Working temperature: -20℃ -70℃

Storage environment: -40℃ -125 ℃, <90%RH

 

[ESP32-CAM]

Modle: ESP32-CAM

Operating Voltage: 5V

SPI Flash: 32 Mbit

RAM: Inter 520KB+, External 4M PSRAM

Bluetooth: BLuetooth 4.2 BR/EDR & BLE standard

WIFI: 802.11/b/g/n/e/i

Port: UART, SPI, I2C, PWM

IO Port: 9

Serial port rate: 115200 bps

Picture form: JPEG (only OV2640 support), BMP, GRAYSCALE

Spectrum range: 2412 -2484MHz

Antanna: Onboard PCB Antanna 2dBi

Security: WPA/WPA2/WPA2-Enterprise/WPS

Operating Temperature: -20 -85℃

Storage Tepmerature: -40 -90℃, <90%RH

 

Transmit power: 

802.11b: 17+/-2 dBm (@11Mbps)

802.11g: 14+/-2 dBm (@54Mbps)

802.11n: 13+/-2 dBm (@MSC7)

 

Receive Sensitivity:

CCK, 1 Mbps: -90dBm

CCK, 11 Mbps: -85dBm

6 Mbps(1/2 BPSK): -88dBm

54 Mbps (3/4 64-QAM): -70dBm

MCS7 (65 Mbps, 72.2 Mbps): -67dBm

 

[ESP32-S]

The main chip uses a low-power dual-core 32-bit CPU with a frequency of up to 240MHz and a computing power of up to 600DMIPS.

Default 32Mbit SPI Flash, 520KB SRAM

Support SoftAP and Station mode

Ultra-small 802.11b/g/n Wi-Fi + BT/BLE SoC module

Support UART/SPI/I2C/I2S/PWM/ADC/DAC, etc.

Support for firmware upgrade (FOTA)

Antenna supports on-board antenna or IPEX block output

 

[OV2640 Camera Module]

Feature:

The OV2640 image sensor features 2 megapixels (1632x1232 pixels)

Small size, low operating voltage, and provides all the features of a single chip UXGA camera and image processor.

Through the SCCB control, you can output the entire frame, sub-sampling, take the window, and a variety of resolution 10-bit sampling data.

The product UXGA images up to 15 frames per second.

The user can fully control the image quality, data format and transmission mode.

 

Specification:

High sensitivity for low-light applications

Low voltage for embedded applications

Standard SCCB interface, compatible with I2C interface

RawRGB, RGB (GRB4: 2: 2, RGB565 / 555/444), YUV (4: 2: 2) and YCbCr (4: 2: 2) output format

Supports UXGA, SXGA, VGA, QVGA, QQVGA, CIF, QCIF and up to 40x30 size

Supports Vario Pixel sub sampling mode

Automatic impact control functions include: automatic exposure control, Automatic gain control, automatic white balance, automatic elimination of light stripes, automatic black level calibration. Image quality control including color saturation, hue, gamma, sharpness ANTI_BLOOM

ISP with noise cancellation and dead pixel compensation

Support image scaling

Lens loss compensation

Saturation automatic adjustment

Edge enhancement automatic adjustment

Noise reduction automatic adjustment

Sensing array 1632 x 1232

Maximum format UXGA IO

Voltage 1.7V -3.3V

Analog voltage 2.5 -3.0V (internal LDO to power supply 1.2V)

Power consumption TBD sleep<20uA

Temperature operation -30 -70℃

Stable working 0 to 50 Degrees Celsius

Output format (8-bit)

YUV / YCbCr4: 2: 2 RGB565 / 555/444

GRB4: 2: 2 Raw RGB Data

Optical size 1/4"

Viewing angle 25°

Maximum rate 15fps

SXGA sensitivity 1.3V / (Lux-sec )

Signal to noise ratio 40dB

Dynamic range 50 dB

View mode Progressive electron exposure 1 line to 1247 lines

Pixel area 2.2 um x 2.2 um

Dark current 15 mV / s at 60℃

ESP32-Cam-tank: fpv vehicle powered by esp32 cam

Kun Li created this amazing Tank Teleoperated Robot. 

ESP32-Cam-tank

a fpv vehicle powered by esp32 cam.
This project use a esp32-cam chip to make a common toy into a fpv rc toy.
It can record videos. It also has telemetry to monitoring the battery voltage and wifi signal strength.
Welcome to modify/upgrade and share.

material list:
ESP32-cam x 1
MX612/MX612E x 2
dc-dc converter 1.5~4.2v to 5.2v x 1
resistor a 5.1kΩ x 1
resistor b 3.2kΩ x 1
capacitor a 100μf x 1
capacitor b 10μf x 2
3.7v lithium battery x 1
switch x 1
tank base with motor and gear x 1
wire x n

Find the code here. 

https://github.com/leoncoolmoon/ESP32-Cam-tank