deepstream_yolo/docs/YOLOv7.md

YOLOv7 usage

NOTE: The yaml file is not required.

• Convert model
• Compile the lib
• Edit the config_infer_primary_yoloV7 file
• Edit the deepstream_app_config file
• Testing the model

Convert model

1. Download the YOLOv7 repo and install the requirements

git clone https://github.com/WongKinYiu/yolov7.git
cd yolov7
pip3 install -r requirements.txt
pip3 install onnx onnxsim onnxruntime

NOTE: It is recommended to use Python virtualenv.

2. Copy conversor

Copy the export_yoloV7.py file from DeepStream-Yolo/utils directory to the yolov7 folder.

3. Download the model

Download the pt file from YOLOv7 releases (example for YOLOv7)

wget https://github.com/WongKinYiu/yolov7/releases/download/v0.1/yolov7.pt

NOTE: You can use your custom model.

4. Reparameterize your model

YOLOv7 and its variants cannot be directly converted to engine file. Therefore, you will have to reparameterize your model using the code here. Make sure to convert your custom checkpoints in yolov7 repository, and then save your reparmeterized checkpoints for conversion in the next step.

5. Convert model

Generate the ONNX model file (example for YOLOv7)

python3 export_yoloV7.py -w yolov7.pt --dynamic

NOTE: To simplify the ONNX model (DeepStream >= 6)

--simplify

NOTE: To use dynamic batch-size (DeepStream >= 6)

--dynamic

NOTE: To use implicit batch-size (example for batch-size = 4)

--batch 4

NOTE: If you are using DeepStream 5.1, remove the --dynamic arg and use opset 12 or lower. The default opset is 12.

--opset 12

NOTE: To convert a P6 model

--p6

NOTE: To change the inference size (defaut: 640)

-s SIZE
--size SIZE
-s HEIGHT WIDTH
--size HEIGHT WIDTH

Example for 1280

-s 1280

or

-s 1280 1280

6. Copy generated files

Copy the generated ONNX model file and labels.txt file (if generated) to the DeepStream-Yolo folder.

Compile the lib

Open the DeepStream-Yolo folder and compile the lib

• DeepStream 6.2 on x86 platform

  CUDA_VER=11.8 make -C nvdsinfer_custom_impl_Yolo
  

• DeepStream 6.1.1 on x86 platform

  CUDA_VER=11.7 make -C nvdsinfer_custom_impl_Yolo
  

• DeepStream 6.1 on x86 platform

  CUDA_VER=11.6 make -C nvdsinfer_custom_impl_Yolo
  

• DeepStream 6.0.1 / 6.0 on x86 platform

  CUDA_VER=11.4 make -C nvdsinfer_custom_impl_Yolo
  

• DeepStream 5.1 on x86 platform

  CUDA_VER=11.1 make -C nvdsinfer_custom_impl_Yolo
  

• DeepStream 6.2 / 6.1.1 / 6.1 on Jetson platform

  CUDA_VER=11.4 make -C nvdsinfer_custom_impl_Yolo
  

• DeepStream 6.0.1 / 6.0 / 5.1 on Jetson platform

  CUDA_VER=10.2 make -C nvdsinfer_custom_impl_Yolo
  

Edit the config_infer_primary_yoloV7 file

Edit the config_infer_primary_yoloV7.txt file according to your model (example for YOLOv7 with 80 classes)

[property]
...
onnx-file=yolov7.onnx
...
num-detected-classes=80
...
parse-bbox-func-name=NvDsInferParseYolo
...

NOTE: The YOLOv7 resizes the input with center padding. To get better accuracy, use

...
maintain-aspect-ratio=1
symmetric-padding=1
...

NOTE: By default, the dynamic batch-size is set. To use implicit batch-size, uncomment the line

...
force-implicit-batch-dim=1
...

Edit the deepstream_app_config file

...
[primary-gie]
...
config-file=config_infer_primary_yoloV7.txt

Testing the model

deepstream-app -c deepstream_app_config.txt

NOTE: The TensorRT engine file may take a very long time to generate (sometimes more than 10 minutes).

NOTE: For more information about custom models configuration (batch-size, network-mode, etc), please check the docs/customModels.md file.