cv-detectionlisted

## Object Detection Best Practice Architecture families: - One-stage: YOLO (v5/v8), SSD, RetinaNet, FCOS - Two-stage: Faster R-CNN, Cascade R-CNN - Transformer: DETR, DINO, RT-DETR Training recipe: - Use pre-trained backbone (ImageNet) - Multi-scale training and testing - IoU threshold: 0.5 for mAP50, 0.5:0.95 for mAP - Use FPN for multi-scale feature extraction - Focal loss for class imbalance in one-stage detectors Standard benchmarks: - COCO val2017: ~37 mAP (Faster R-CNN R50), ~51 mAP (DINO Swin-L) - Pascal VOC: ~80 mAP50 (Faster R-CNN)