EfficientNet-V2-s 示例

此文档讲述使用 QAI AppBuilder Python API 利用 Qualcomm® Hexagon™ Processor (NPU) 推理 EfficientNet-V2-s 目标识别模型。

示例支持设备

设备	SoC
Dragon Q6A	QCS6490
Dragon Q8B	SC8280XP

安装 QAI AppBuilder

提示

1. 请根据 QAI AppBuilder 安装方法 安装 QAI AppBuilder。

2. 请根据 创建 ADSP 环境变量 配置 ADSP 环境变量。

运行示例

安装依赖

Device

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

运行脚本

• 进入示例目录

  QCS6490

  Device

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

  SC8280XP

  Device

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

• 准备输入图片，这里以以下图片为输入示例

  

  input image

• 执行推理

  Device

  python3 efficientnet_v2_s/efficientnet_v2_s.py

  $ python3 efficientnet_v2_s/efficientnet_v2_s.py
  Current file directory: /mnt/ssd/qualcomm/701/zzf_fork/ai-engine-direct-helper/samples/linux/python/efficientnet_v2_s
       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
  
     129.0ms [WARNING] Time: Read model file to memory. 11.18
  
       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
  
     158.9ms [WARNING] Time: contextCreateFromBinary. 29.89
  
     159.0ms [WARNING] Time: UnmapViewOfFile. 0.00
  
     159.4ms [WARNING] Time: model_initialize efficientnet_v2_s 159.34
  
     227.4ms [WARNING] Time: model_inference efficientnet_v2_s 10.06
  
  Top 5 predictions for image:
  
  Samoyed 0.7607400417
  keeshond 0.0230328161
  Pomeranian 0.0204458293
  white wolf 0.0193288065
  Arctic fox 0.0062915389
       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
     239.8ms [WARNING] Time: model_destroy efficientnet_v2_s 9.40

  打印结果可以看到 Samoyed 置信度最高，这与输入图片内容吻合。