[🚀🚀🚀CoreML Yolo🚀🚀🚀]

中文README

A unified C++ and Objective-C++ toolkit for YOLO v5/v8/v11/v26/..., covering classification/detection/segmentation/pose/obb tasks with easy python-like APIs from ultralytics/ultralytics. Support All Yolo Tasks, All Yolo Versions, All Yolo Runtimes, it's time to make all in one.

1. demo code supprots Yolo tasks like detection/ obb.
2. support all Yolo versions including yolov5(anchor-based)/yolov5u(anchor-free)/yolov8/yolov11/yolov26(nms-free)/more in the future, sub versions like n/s/m/l/x are also supported.
3. supports CoreML Yolo backends(runtime)
4. easy APIs to use and integrate, as simple as python APIs from ultralytics/ultralytics library.
5. toolkit works out of box, provide the model and set up the config parameters, go predict!

It is tested on Apple Silicon macs with 64GB UM on 6 and 8way video running 54 and 42 fps avg.

1. C++ >= 17
2. GCC >= 7.5 or clang
3. OpenCV == 4.13
4. CUDA/ONNXRuntime/TensorRT/OpenVINO/RKNN/... are optional

1. run git clone https://github.com/edselmalasig/one-yolo-coreml.git
2. run cd one-yolo && mkdir build-macos && cd build-macos
3. run cmake .. && make -j8

build options when run cmake command:
-DBUILD_WITH_CML=ON   # enable CoreML(Apple Platform) as inference backend
-DBUILD_WITH_PDL=ON   # enable PaddlePaddle as inference backend
-DBUILD_WITH_CAN=ON   # enable CANN(HuaWei Platform) as inference backend
-DBUILD_WITH_DEL=ON   # enable Denglin's SDK(Denglin/登临 Platform) as inference backend
-DBUILD_WITH_CAB=ON   # enable Cambricon'SDK(Cambricon/寒武纪 Platform) as inference backend
...

if you just run `cmake ..` without any options, 
one-yolo will depend on OpenCV::DNN module as inference backend by default,
so OpenCV is required for one-yolo, CUDA is optional when building OpenCV from source code. 

you can get yolo models .pt from http://www.ultralytics.com

1. pip install ultralytics
2. yolo export model=yolo26n.pt format=coreml
3. copy the assets folder to build-macos/samples
4. cd build-macos/samples
5. mkdir models
6. move the converted .pt package... the .mlpackage to the models directory
7. run the executable from the terminal

can be found in assets folder