Support extensive shape for functions related to bounding box localization

[developer0hye]

This PR will make us play with bounding boxes more easily.

Signed-off-by: Yonghye Kwon developer.0hye@gmail.com

🛠️ PR Summary

_{Made with ❤️ by Ultralytics Actions}

🌟 Summary

Enhancement of bounding box coordinate transformations to support higher dimensional tensors.

📊 Key Changes

• Implemented the use of ellipsis (...) in array indexing for box conversion functions (e.g., xyxy2xywh, xywh2xyxy, etc.), which allows these functions to operate on inputs with arbitrary number of dimensions.
• The changes apply to both PyTorch tensor and NumPy array operations, maintaining functionality across different object types.
• Adjusted clip_boxes function to support clipping for higher-dimensional tensors utilizing the same ellipsis indexing pattern.

🎯 Purpose & Impact

• Flexibility: The modifications introduce flexibility in handling bounding boxes, enabling batch processing and multi-dimensional data, leading to cleaner code when dealing with datasets that contain multiple images or annotations.
• Performance: Batch operations can improve performance, allowing more efficient computations when processing images or video sequences in a machine learning context, specifically within object detection tasks.
• Consistency: All related functions now follow the same indexing pattern, increasing the maintainability of the codebase.
• Impact: This will be particularly beneficial for users and developers working on applications involving a high volume of data, such as video analysis or large-scale image processing, ultimately enhancing the YOLOv5's usability in complex machine learning pipelines. 🚀

[glenn-jocher]

@developer0hye PR is merged. Thank you for your contributions to YOLOv5 🚀 and Vision AI ⭐

[developer0hye]

@glenn-jocher

Hi Jocher!
Is there way to talk with you on a messenger?

[glenn-jocher]

@developer0hye yes, you can message me on LinkedIn at https://www.linkedin.com/in/glenn-jocher/ or WhatsApp at +13012376695

[developer0hye]

@glenn-jocher Okay, thanks!

[glenn-jocher]

@developer0hye you're welcome! If you have any more questions or need assistance, feel free to ask here on GitHub. We're always happy to help 😊