kaiven11/coreml_for_ios

Deploy CoreML on IOS for object detection

supporting models: (https://github.com/ultralytics/yolov5), [YOLOv5]

This project is Object Detection on iOS with Core ML.

Requirements

• Xcode 14.x
• iOS 16.0+
• Swift 5.7.x

How To Build and Run the Project

1. Clone the project

git clone https://github.com/kaiven11/coreml_for_ios

2. Add the model to the project

Rename the ObjectDetection-CoreML.PY file to ObjectDetection-CoreML.xcodeproj,and use Xcode to open it. By default, the project uses the yolov5s model. you can use export-nms.py to export pt to CoreML file.

python export-nms.py --include coreml --weights yolov5n.pt

3. Set model name properly in ViewController.swift

4. Build and Run

See also

• motlabs/awesome-ml-demos-with-ios
  : The challenge using machine learning model created from tensorflow on iOS : YOLOv8 repository
• ultralytics/yolov5
  : YOLOv5 repository