MCUXpresso_MIMXRT1052xxxxB/boards/evkbimxrt1050/eiq_examples/glow_cifar10_camera

Overview

Demonstrates inference for models compiled using the GLOW AOT tool and uses a camera to generate data for inferencing.

The network used in this is based on the CIFAR-10 example in Caffe2 [1] & [2].

[1] https://github.com/caffe2/tutorials/blob/master/CIFAR10_Part1.ipynb [2] https://github.com/caffe2/tutorials/blob/master/CIFAR10_Part2.ipynb

The neural network consists of 3 convolution layers interspersed by ReLU activation and max pooling layers, followed by a fully-connected layer at the end. The input to the network is a 32x32 pixel color image extracted from camera data, which will be classified into one of the 10 output classes.

This project does not include the pre-trained model or the training script since Caffe2 framework is deprecated and lately has become part of PyTorch. This project example only includes the bundle (binary) generated after running the Glow AOT tool and is intended to be used as-is. The Glow bundle is the same as the "glow_cifar10" MCUXpresso SDK example in the SDK. See that project's readme.txt for more details on model generation.

If you want a step-by-step example of running the Glow AOT tool for a given model take a look at the LeNet MNIST Glow example and the Glow Getting Started Lab: https://community.nxp.com/t5/eIQ-Machine-Learning-Software/eIQ-Glow-Lab-for-i-MX-RT/ta-p/1123119

Toolchains supported

• MCUXpresso IDE
• IAR Embedded Workbench for ARM
• Keil uVision MDK
• ArmGCC - GNU Tools ARM Embedded

SDK version

• Version: 2.15.000

Toolchain supported

• IAR embedded Workbench 9.40.1
• Keil MDK 5.38.1
• GCC ARM Embedded 12.2
• MCUXpresso 11.8.0

Hardware requirements

• Mini/micro USB cable
• EVKB-IMXRT1050 board
• Personal computer
• MT9M114 camera (optional)
• RK043FN02H-CT display (optional)

Board settings

Connect the camera to J35 (optional) Connect the display to A1-A40 and B1-B6 (optional) Connect external 5V power supply to J2, set J1 to 1-2

Prepare the Demo

1. Connect a USB cable between the host PC and the OpenSDA USB port on the target board.
2. Open a serial terminal with the following settings:
   • 115200 baud rate
   • 8 data bits
   • No parity
   • One stop bit
   • No flow control
3. Download the program to the target board.
4. Either press the reset button on your board or launch the debugger in your IDE to begin running the demo.

Running the demo

Use the LCD screen to point the camera at images of airplanes, ships, deer, frogs, cars, and other images that can be categorized by CIFAR10. Some images will work better than others, and a few example images have been provided in the PDF in the /doc folder. For best results, the flashing selection rectangle should be centered on the image and nearly (but not completely) fill up the whole rectangle. The camera should be stabilized with your finger or by some other means to prevent shaking. Also ensure the camera lens has been focused as described in the instructions when connecting the camera and LCD (https://community.nxp.com/t5/i-MX-RT-Knowledge-Base/Connecting-camera-and-LCD-to-i-MX-RT-EVKs/tac-p/1122184).

You will see the result of the inference on the LCD screen as well as the serial terminal. The result printed on the LCD screen has a minimum threshold applied to it.