Sample Overview

This document uses a simple image classification application as an example to describe how to use mxVision to orchestrate an application. As shown in Figure 1, data is first received, then the ResNet-50 model is used to classify the images, and finally the classification result is displayed.

Figure 1 Image classification application

Introduction to the ResNet-50 model:

• Input data: RGB input images with 224 x 224 resolution.
• Output data: class label and confidence value of each image. (The confidence value indicates the possibility that an image belongs to a certain class.)