ConvNext-Base Example

This document describes using the QAI AppBuilder Python API to perform inference with the ConvNext-Base target recognition model using Qualcomm® Hexagon™ Processor (NPU).

Supported Devices

Device	SoC
Dragon Q6A	QCS6490
Dragon Q8B	SC8280XP

Install QAI AppBuilder

tip

1. Please install QAI AppBuilder according to QAI AppBuilder Installation Guide.

2. Please configure QAIRT environment variables according to Configure QAIRT Environment Variables.

Run Example

Install Dependencies

Device

pip3 install requests tqdm qai-hub py3_wget opencv-python torch torchvision

Run Script

• Navigate to the example directory

  QCS6490

  Device

  cd ai-engine-direct-helper/samples/linux/python

  SC8280XP

  Device

  cd ai-engine-direct-helper/samples/linux/python

• Prepare input image, using the following image as input example

  

  input image

• Execute inference

  Device

  python3 convnext_base/convnext_base.py

  $ python3 convnext_base/convnext_base.py
  Current file directory: /mnt/ssd/qualcomm/701/zzf_fork/ai-engine-direct-helper/samples/linux/python/convnext_base
       0.0ms [WARNING]  <W> Initializing HtpProvider
  
  /prj/qct/webtech_scratch20/mlg_user_admin/qaisw_source_repo/rel/qairt-2.37.1/point_release/SNPE_SRC/avante-tools/prebuilt/dsp/hexagon-sdk-5.4.0/ipc/fastrpc/rpcmem/src/rpcmem_android.c:38:dummy call to rpcmem_init, rpcmem APIs will be used from libxdsprpc
       0.0ms [WARNING]  <W> This META does not have Alloc2 Support
  
       0.0ms [WARNING]  <W> This META does not have Alloc2 Support
  
       0.0ms [WARNING]  <W> This META does not have Alloc2 Support
  
       0.0ms [WARNING]  <W> This META does not have Alloc2 Support
  
     193.9ms [WARNING] Time: Read model file to memory. 71.69
  
       0.0ms [WARNING]  <W> This META does not have Alloc2 Support
  
       0.0ms [WARNING]  <W> This META does not have Alloc2 Support
  
       0.0ms [WARNING]  <W> This META does not have Alloc2 Support
  
       0.0ms [WARNING]  <W> This META does not have Alloc2 Support
  
       0.0ms [WARNING]  <W> This META does not have Alloc2 Support
  
       0.0ms [WARNING]  <W> This META does not have Alloc2 Support
  
       0.0ms [WARNING]  <W> This META does not have Alloc2 Support
  
     269.5ms [WARNING] Time: contextCreateFromBinary. 75.47
  
     269.5ms [WARNING] Time: UnmapViewOfFile. 0.00
  
     270.9ms [WARNING] Time: model_initialize convnext_base 270.85
  
     348.3ms [WARNING] Time: model_inference convnext_base 30.17
  
  Top 5 predictions for image:
  
  Samoyed 0.8726790547
  Pomeranian 0.0192712061
  keeshond 0.0150661934
  Japanese spaniel 0.0032834315
  Eskimo dog 0.0020474677
       0.0ms [WARNING]  <W> This META does not have Alloc2 Support
  
       0.0ms [WARNING]  <W> This META does not have Alloc2 Support
  
       0.0ms [WARNING]  <W> This META does not have Alloc2 Support
  
       0.0ms [WARNING]  <W> This META does not have Alloc2 Support
  
  /prj/qct/webtech_scratch20/mlg_user_admin/qaisw_source_repo/rel/qairt-2.37.1/point_release/SNPE_SRC/avante-tools/prebuilt/dsp/hexagon-sdk-5.4.0/ipc/fastrpc/rpcmem/src/rpcmem_android.c:42:dummy call to rpcmem_deinit, rpcmem APIs will be used from libxdsprpc
     365.5ms [WARNING] Time: model_destroy convnext_base 14.44

  The printed results show that Samoyed has the highest confidence, which matches the input image content.