Skip to content

Latest commit

 

History

History
247 lines (160 loc) · 6.9 KB

README.md

File metadata and controls

247 lines (160 loc) · 6.9 KB
Raw

yolov5

Table of contents

1. Description

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

https://github.com/airockchip/yolov5

2. Current Support Platform

RK3566, RK3568, RK3588, RK3562, RV1106, RV1103, RV1109, RK3576, RV1126, RK1808, RK3399PRO

3. Pretrained Model

Download link:

./yolov5s_relu.onnx
./yolov5n.onnx
./yolov5s.onnx
./yolov5m.onnx

Download with shell command:

cd model
./download_model.sh

Note: The model provided here is an optimized model, which is different from the official original model. Take yolov5n.onnx as an example to show the difference between them.

  1. The comparison of their output information is as follows. The left is the official original model, and the right is the optimized model.
Image
  1. Taking the output change [1,19200,85]->[1,255,80,80] as an example, we delete a subgraph (the framed part in the picture) from the model and put it in post-processing (this subgraph is not quantification-friendly)
Image

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/yolov5s_relu.onnx rk3588
# output model will be saved as ../model/yolov5.rknn

Description:

  • <onnx_model>: Specify ONNX model path.
  • <TARGET_PLATFORM>: Specify NPU platform name. Support Platform refer here.
  • <dtype>(optional): Specify as i8 or fp. i8 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 yolov5.rknn

5. Python Demo

Usage:

cd python
# Inference with PyTorch model or ONNX model
python yolov5.py --model_path <pt_model/onnx_model> --img_show

# Inference with RKNN model
python yolov5.py --model_path <rknn_model> --target <TARGET_PLATFORM> --img_show

Description:

  • <TARGET_PLATFORM>: Specify NPU platform name. Such as 'rk3588'.

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

6. Android Demo

6.1 Compile and Build

Usage:

# go back to the rknn_model_zoo root directory
cd ../../
export ANDROID_NDK_PATH=<android_ndk_path>

./build-android.sh -t <TARGET_PLATFORM> -a <ARCH> -d yolov5

# such as 
./build-android.sh -t rk3588 -a arm64-v8a -d yolov5

Description:

  • <android_ndk_path>: Specify Android NDK path.
  • <TARGET_PLATFORM>: Specify NPU platform name. Support Platform refer here.
  • <ARCH>: Specify device system architecture. To query device architecture, refer to the following command: 
     # Query architecture. For Android, ['arm64-v8a' or 'armeabi-v7a'] should shown in log.
 adb shell cat /proc/version

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_yolov5_demo/ /data/

6.3 Run demo

adb shell
cd /data/rknn_yolov5_demo

export LD_LIBRARY_PATH=./lib
./rknn_yolov5_demo model/yolov5.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_yolov5_demo/out.png

7. Linux Demo

7.1 Compile and Build

usage

# go back to the rknn_model_zoo root directory
cd ../../

# if GCC_COMPILER not found while building, please set GCC_COMPILER path
(optional)export GCC_COMPILER=<GCC_COMPILER_PATH>

./build-linux.sh -t <TARGET_PLATFORM> -a <ARCH> -d yolov5

# such as 
./build-linux.sh -t rk3588 -a aarch64 -d yolov5
# such as 
./build-linux.sh -t rv1106 -a armhf -d yolov5

Description:

  • <GCC_COMPILER_PATH>: Specified as GCC_COMPILER path.

    • For RV1106, RV1103, GCC_COMPILER version is arm-rockchip830-linux-uclibcgnueabihf 
      export GCC_COMPILER=~/opt/arm-rockchip830-linux-uclibcgnueabihf/bin/arm-rockchip830-linux-uclibcgnueabihf

  • <TARGET_PLATFORM> : Specify NPU platform name. Support Platform refer here.

  • <ARCH>: Specify device system architecture. To query device architecture, refer to the following command:

    # Query architecture. For Linux, ['aarch64' or 'armhf'] should shown in log.
adb shell cat /proc/version

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_yolov5_demo/ /userdata/
  • For other boards, use scp or other approaches to push all files under install/<TARGET_PLATFORM>_linux_<ARCH>/rknn_yolov5_demo/ to userdata.

7.3 Run demo

adb shell
cd /userdata/rknn_yolov5_demo

export LD_LIBRARY_PATH=./lib
./rknn_yolov5_demo model/yolov5.rknn model/bus.jpg

  • RV1106/1103 LD_LIBRARY_PATH must specify as the absolute path. Such as

    export LD_LIBRARY_PATH=/userdata/rknn_yolov5_demo/lib

  • 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_yolov5_demo/out.png

8. Expected Results

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

person @ (209 244 286 506) 0.884
person @ (478 238 559 526) 0.868
person @ (110 238 230 534) 0.825
bus @ (94 129 553 468) 0.705
person @ (79 354 122 516) 0.339


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