Skip to content

Latest commit

 

History

History
185 lines (113 loc) · 4.99 KB

README.md

File metadata and controls

185 lines (113 loc) · 4.99 KB

ppyoloe

Table of contents

1. Description

The model used in this example comes from the following open source projects:

https://github.com/PaddlePaddle/PaddleDetection/blob/release/2.6/configs/ppyoloe

  • To export the model matching this demo requirement, refer here to export ONNX model.

2. Current Support Platform

RK3566, RK3568, RK3588, RK3562, RK1808, RV1109, RV1126

3. Pretrained Model

Download link:

./ppyoloe_s.onnx
./ppyoloe_m.onnx

Download with shell command:

cd model
./download_model.sh

4. Convert to RKNN

Usage:

cd python
python convert.py <onnx_model> <TARGET_PLATFORM> <dtype(optional)> <output_rknn_path(optional)>

# such as: 
python convert.py ../model/ppyoloe_s.onnx rk3588
# output model will be saved as ../model/ppyoloe.rknn

Description:

  • <onnx_model>: Specify ONNX model path.
  • <TARGET_PLATFORM>: Specify NPU platform name. Support Platform refer [here](#2 Current Support Platform).
  • <dtype>(optional): Specify as i8, u8 or fp. i8/u8 for doing quantization, fp for no quantization. Default is i8.
  • <output_rknn_path>(optional): Specify save path for the RKNN model, default save in the same directory as ONNX model with name ppyoloe.rknn

5. Python Demo

Usage:

cd python
# Inference with PyTorch model or ONNX model
python ppyoloe.py --model_path <pt_model/onnx_model> --img_show

# Inference with RKNN model
python ppyoloe.py --model_path <rknn_model> --target <TARGET_PLATFORM> --img_show

Description:

  • <TARGET_PLATFORM>: Specify NPU platform name. Support Platform refer [here](#2 Current Support Platform).

  • <pt_model / onnx_model / rknn_model>: Specify the model path.

6. Android Demo

Note: RK1808, RV1109, RV1126 does not support Android.

6.1 Compile and Build

Please refer to the Compilation_Environment_Setup_Guide document to setup a cross-compilation environment and complete the compilation of C/C++ Demo.
Note: Please replace the model name with ppyoloe.

6.2 Push demo files to device

With device connected via USB port, push demo files to devices:

adb root
adb remount
adb push install/<TARGET_PLATFORM>_android_<ARCH>/rknn_ppyoloe_demo/ /data/

6.3 Run demo

adb shell
cd /data/rknn_ppyoloe_demo

export LD_LIBRARY_PATH=./lib
./rknn_ppyoloe_demo model/ppyoloe.rknn model/bus.jpg

  • After running, the result was saved as out.png. To check the result on host PC, pull back result referring to the following command:

    adb pull /data/rknn_ppyoloe_demo/out.png

7. Linux Demo

7.1 Compile and Build

Please refer to the Compilation_Environment_Setup_Guide document to setup a cross-compilation environment and complete the compilation of C/C++ Demo.
Note: Please replace the model name with ppyoloe.

7.2 Push demo files to device

  • If device connected via USB port, push demo files to devices:
adb push install/<TARGET_PLATFORM>_linux_<ARCH>/rknn_ppyoloe_demo/ /userdata/
  • For other boards, use scp or other approaches to push all files under install/<TARGET_PLATFORM>_linux_<ARCH>/rknn_ppyoloe_demo/ to userdata.

7.3 Run demo

adb shell
cd /userdata/rknn_ppyoloe_demo

export LD_LIBRARY_PATH=./lib
./rknn_ppyoloe_demo model/ppyoloe.rknn model/bus.jpg

  • After running, the result was saved as out.png. To check the result on host PC, pull back result referring to the following command:

    adb pull /userdata/rknn_ppyoloe_demo/out.png

8. Expected Results

This example will print the labels and corresponding scores of the test image detect results, as follows:

person @ (108 234 225 536) 0.947
person @ (477 233 561 519) 0.936
person @ (211 240 283 512) 0.925
bus @ (86 135 552 445) 0.921
person @ (79 326 124 516) 0.489
handbag @ (260 338 281 413) 0.406
person @ (103 243 116 285) 0.253

  • Note: Different platforms, different versions of tools and drivers may have slightly different results.