Log

[john-rocky]

Added a log function for developer mode.

sec_day, free_space, battery_level ,id, confidence, box_x, box_y, box_w, box_h, weight, height, age, gender, gender_confidence, race, race_confidence are recorded in a csv file for each object detected in each frame. In detect mode, only box is recorded.

When you turn on developer mode for Yolo-iOS-app from the device settings app, the save button will appear on the toolbar at the bottom of the app.

Pressing the save button will start recording the log, and pressing the save button again will pop up the activity controller, allowing you to attach the csv file to an email or save it to the file app.

The csv file name is in the format of

human_results_20240626_10.18.03.csv.

The items recorded are based on the sec_day and other items that were originally in the app. Please let us know if you want to change them.

🛠️ PR Summary

_{Made with ❤️ by Ultralytics Actions}

🌟 Summary

This PR updates the ultralytics/yolo-ios-app project with multiple changes including orientation handling, new features for human model tracking, and improvements in the README for better setup instructions.

📊 Key Changes

• README Improvements: Enhanced instructions for setting up the project in Xcode and exporting YOLO models.
• File Additions: Added new Swift files like HumanModel.swift, PostProcessing.swift, TrackingModel.swift, and SaveResults.swift for advanced functionalities.
• Orientation Handling: Improved handling of device orientation changes during video capture.
• UI Updates: Adjusted storyboard for better interface orientation support and added elements for new features.
• Code Enhancements: Improved bounding box handling, added methods for saving detection results, and tracking human models.

🎯 Purpose & Impact

• Enhanced User Experience: By handling device orientation changes and improving the setup instructions, the app becomes easier to use and more intuitive. 📱
• Advanced Features: The introduction of human model tracking and post-processing enhances the app’s functionality, making it suitable for more complex applications. 🏃‍♂️
• Developer Tools: Adding methods to save detection results helps developers debug and analyze model performance more effectively. 💾

These updates collectively make the YOLO iOS app more robust, user-friendly, and feature-rich, facilitating better performance and ease of use for both end-users and developers.