ImageClassification

Image Classification Python CLI Application

Table of contents

• Overview
• Requirements
• Preparing model assets
• Running Python app

1. Overview

Image Classification application for Windows on Snapdragon® with ResNet101 using ONNX runtime.

ResNet101 is a machine learning model that can classify images from the Imagenet dataset. It can also be used as a backbone in building more complex models for specific use cases. Optimized for Qualcomm Compute platform, this application enables real-time image classification.

This Python application demonstrates how to use QNN Execution Provider to accelerate the model using the Snapdragon® Neural Processing Unit (NPU).

2. Requirements

Platform

• Windows on Snapdragon® (Qualcommm Compute platform, e.g. X Elite and X Plus)
• Windows 11
• This application is tested on ASUS Vivobook S15 (S5507).

Tools and SDK

• Python

  • This application is tested with Python 3.10.9.
  • Download and install Python 3.10.9 64-bit on your Windows on Snapdragon.
  • Required packages.
    • numpy
    • onnxruntime-qnn
    • pillow

• Qualcomm AI Runtime SDK : QNN SDK

  • The required QNN dependency libraries are included in onnxruntime-qnn package.
  • If you plan to use a specific version of QNN libraries, download and install Qualcomm AI Runtime SDK from Qualcomm Software Center.
  • This Python application is tested with default QNN libraries from onnxruntime-qnn and QNN v2.39.0.250926.
  • Find your QNN_SDK_ROOT. For example, QNN_SDK_ROOT = C:\Qualcomm\AIStack\QAIRT\2.39.0.250926.
  • Remember this directory if you plan to use a specific version of QNN libraries.
    • <QNN_SDK_ROOT>\lib\arm64x-windows-msvc

3. Preparing model assets

Downloading the model from Qualcomm AI Hub

Download ResNet101 ONNX model for Compute from Qualcomm AI Hub. Extract the zip file and place model files into ./assets/ directory.

Image assets

Prepare your test image assets and place into ./assets/ directory.

Model input resolution is 224 x 224.

If input image is in different resolution, it will be resized to 224 x 224.

4. Running Python app

Checking the assets directory

Please ensure that you have followed the section above and placed the following assets into the specific directory. You may change the directory if needed.

• ONNX model from Qualcomm AI Hub and test images : ./assets/

Installing required Python packages

Open your terminal and navigate to the project directory.

pip install numpy onnxruntime-qnn pillow

Running image classification app via CLI

Run the application with default QNN libraries from onnxruntime-qnn package.

You may change the path if needed.

python .\Image_Classification_ONNX.py --onnx_path .\assets\model.onnx

You can also run the application with a specific version of QNN libraries.

python .\Image_Classification_ONNX.py --onnx_path .\assets\model.onnx --qnn_path C:\Qualcomm\AIStack\QAIRT\2.39.0.250926\lib\arm64x-windows-msvc\QnnHtp.dll

Input your test image path

Enter the path to the image file (or type 'exit' to quit): ./assets/keyboard.jpg

Example Output