Enhancing YOLOv11 for Real-Time Object Detection: Advanced Architectures and Edge-Optimized Training Pipeline

In this paper, we propose novel enhancements to YOLOv11, leveraging its advanced architectural components such as the C3k2 block, SPPF (Spatial Pyramid Pooling -Fast), and C2PSA (Convolutional Block with Parallel Spatial Attention).These innovations address key challenges in real-time object detection, including feature extraction, attention mechanisms, and computational efficiency.Furthermore, we present a new training pipeline that optimizes YOLOv11 for edge computing while maintaining state-of-the-art accuracy.Experimental results on the COCO dataset demonstrate significant improvements in mean Average Precision (mAP) and latency compared to prior YOLO iterations, establishing YOLOv11 as a benchmark for real-time applications.