[NeurIPS 2025] YOLOv12: Attention-Centric Real-Time Object Detectors
All-in-one training for vision models (YOLO, ViTs, RT-DETR, DINOv3): pretraining, fine-tuning, distillation.
在 GitHub 打开