New Release: Ultralytics v8.3.24

:rocket: Ultralytics Release v8.3.24 Announcement

:star2: Summary

We’re excited to announce the release of Ultralytics v8.3.24! This update brings significant enhancements to YOLO’s prediction module, improved documentation for NVIDIA Jetson deployments, and crucial compatibility fixes for macOS users. Dive into the details below to see how these changes can benefit your projects.

:bar_chart: Key Changes

New Features

  • SAM Predict Box Enhancement: The postprocess function now handles predictions more robustly, ensuring default bounding boxes are correctly initialized even when no masks are detected.

Improvements

  • Improved Documentation: Our NVIDIA Jetson guide has been updated from YOLOv8 to YOLO11, providing clearer instructions for better deployment.
  • macOS Compatibility: We’ve restricted the numpy version to address compatibility issues with OpenVINO, enhancing stability for macOS users.
  • CI/CD Optimization: Updates to GitHub Actions improve disk cleanup efficiency and streamline CI trigger conditions.

:dart: Purpose & Impact

  • Robust Predictions: The SAM update ensures reliable prediction processes, maintaining operational reliability across various applications.
  • Ease of Deployment: The updated Jetson documentation supports users transitioning to YOLO11, leveraging new capabilities for AI tasks.
  • Platform Stability: Pinning numpy versions reduces issues during model exports on macOS, improving user experience.
  • Efficient Development Processes: Optimized CI/CD workflows lead to faster and more cost-effective software development cycles.

What’s Changed

Full Changelog: View Here

Try It Out!

We encourage you to try out the new release and share your feedback. Your insights are invaluable in helping us improve and deliver the best possible tools for your AI projects.

Release URL: Ultralytics v8.3.24

Thank you for being a part of the Ultralytics community!