Deploying ML Models on Edge Devices


A walk through the Hardware components

We will get to understand two different hardware platforms. The first is the OpenMV Cam H7 and the second is the TinyML kit.
We'll be using these two platforms but the process should be reproducible on any other device with an Arm Cortex.

1. OpenMV Cam H7

The Open MV Cam H7

2. TinyML kit:


Arduino Nano 33 BLE Sense:

Arduino Nano 33 BLE Sense

OV7675 Camera Module

OV7675 Camera Module

Tiny Machine Learning (TinyML) Shield

Tiny Machine Learning (TinyML) Shield

TinyML Kit Setup

To get started with the TinyML kit, follow these steps:

  • Ensure that the Arduino Nano 33 BLE Sense board is securely mounted onto the TinyML shield. Align the pins of the Arduino Nano with the corresponding headers on the shield and gently press them together until they are firmly connected.
  • Locate the OV7675 camera module. Connect it to the TinyML shield by plugging it into the camera header on the shield. The shield is designed to provide the necessary connections, so you won't need to worry about manually wiring the camera module.
  • Connect any other peripherals required for your project, such as USB cables for programming and power supply.
  • Optionally, if you need to power the system independently, you can connect an external power source to the power input pins on the TinyML shield. Ensure that the power source meets the voltage requirements of the Arduino Nano and the camera module.

The Assembled Tiny Machine Learning Kit