Yolov4 on MacBook Pro 16-inch using GPU = AMD Radeon Pro 5500M 8 GB

[KohlerVG]

The title says it, I want to use Yolov4 on macOS using this GPU = AMD Radeon Pro 5500M 8 GB.
Has anyone achieved this?
Thanks,
Kohler

[AlexeyAB]

• Try to run your custom yolov4.cfg / yolov4.weights without convertation by using OpenCV on MacBook https://docs.opencv.org/master/da/d9d/tutorial_dnn_yolo.html
  • How to install https://docs.opencv.org/master/d0/db2/tutorial_macos_install.html
  • C++: https://github.com/opencv/opencv/blob/master/samples/dnn/object_detection.cpp
  • Python: https://github.com/opencv/opencv/blob/master/samples/dnn/object_detection.py

---

• Convert your custom yolov4.cfg / yolov4.weights model to ONNX: https://github.com/Tianxiaomo/pytorch-YOLOv4#3-darknet2onnx
  • install OpenMP brew install libomp
  • install onnx-runtime pip install onnxruntime
  • clone repo git clone https://github.com/Tianxiaomo/pytorch-YOLOv4
  • cd pytorch-YOLOv4
  • run convertation and demo python demo_darknet2onnx.py yolov4.cfg yolov4.weights dog.jpg 0

[KohlerVG]

Thank you so much for the instructions! I'll do it and let you know how it goes! Thanks! Kohler

…

On Jan 22, 2021, at 7:02 PM, Alexey ***@***.***> wrote: Try to run your custom yolov4.cfg / yolov4.weights without convertation by using OpenCV on MacBook https://docs.opencv.org/master/da/d9d/tutorial_dnn_yolo.html <https://docs.opencv.org/master/da/d9d/tutorial_dnn_yolo.html> How to install https://docs.opencv.org/master/d0/db2/tutorial_macos_install.html <https://docs.opencv.org/master/d0/db2/tutorial_macos_install.html> C++: https://github.com/opencv/opencv/blob/master/samples/dnn/object_detection.cpp <https://github.com/opencv/opencv/blob/master/samples/dnn/object_detection.cpp> Python: https://github.com/opencv/opencv/blob/master/samples/dnn/object_detection.py <https://github.com/opencv/opencv/blob/master/samples/dnn/object_detection.py> Convert your custom yolov4.cfg / yolov4.weights model to ONNX: https://github.com/Tianxiaomo/pytorch-YOLOv4#3-darknet2onnx <https://github.com/Tianxiaomo/pytorch-YOLOv4#3-darknet2onnx> install OpenMP brew install libomp install onnx-runtime pip install onnxruntime clone repo git clone https://github.com/Tianxiaomo/pytorch-YOLOv4 cd pytorch-YOLOv4 run convertation and demo python demo_darknet2onnx.py yolov4.cfg yolov4.weights dog.jpg 0 — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#7272 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/APR34KMNNVRVO54TJVFN2QDS3IN2BANCNFSM4WNTX6IQ>.

[Dvalin21]

Thank you so much for the instructions! I'll do it and let you know how it goes! Thanks! Kohler
…
On Jan 22, 2021, at 7:02 PM, Alexey @.***> wrote: Try to run your custom yolov4.cfg / yolov4.weights without convertation by using OpenCV on MacBook https://docs.opencv.org/master/da/d9d/tutorial_dnn_yolo.html https://docs.opencv.org/master/da/d9d/tutorial_dnn_yolo.html How to install https://docs.opencv.org/master/d0/db2/tutorial_macos_install.html https://docs.opencv.org/master/d0/db2/tutorial_macos_install.html C++: https://github.com/opencv/opencv/blob/master/samples/dnn/object_detection.cpp https://github.com/opencv/opencv/blob/master/samples/dnn/object_detection.cpp Python: https://github.com/opencv/opencv/blob/master/samples/dnn/object_detection.py https://github.com/opencv/opencv/blob/master/samples/dnn/object_detection.py Convert your custom yolov4.cfg / yolov4.weights model to ONNX: https://github.com/Tianxiaomo/pytorch-YOLOv4#3-darknet2onnx https://github.com/Tianxiaomo/pytorch-YOLOv4#3-darknet2onnx install OpenMP brew install libomp install onnx-runtime pip install onnxruntime clone repo git clone https://github.com/Tianxiaomo/pytorch-YOLOv4 cd pytorch-YOLOv4 run convertation and demo python demo_darknet2onnx.py yolov4.cfg yolov4.weights dog.jpg 0 — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#7272 (comment)>, or unsubscribe https://github.com/notifications/unsubscribe-auth/APR34KMNNVRVO54TJVFN2QDS3IN2BANCNFSM4WNTX6IQ.

Wanted to see if you were able to achieve this? Thanks.