YOLO Series by Ultralytics
A family of real-time object detection models designed for speed and accuracy across diverse computer vision tasks.
• Includes YOLOv3 through YOLOv11, with continuous advancements in model speed, accuracy, and deployment flexibility
• YOLOv8 introduced a modular architecture, anchor-free detection, and support for classification, segmentation, and pose estimation
• YOLOv9 added an RT-DETR head for transformer-based performance with high FPS and accuracy
• YOLOv10 focused on lightweight deployment and mobile optimization using NAS (Neural Architecture Search) and attention modules
• YOLOv11, the latest release, integrates improved quantization, transformer-based architecture refinements, and multi-task learning for simultaneous detection and segmentation tasks
• Easily trainable on custom datasets with strong out-of-the-box performance
• Extensive support for deployment on edge devices, cloud platforms, and various frameworks (PyTorch, ONNX, TensorRT)
• Open-source core with proprietary deployment tools and cloud integrations available via Ultralytics HUB