Researchers have developed various models for detection and image processing tasks, including defect detection, object detection, and image restoration. These models utilize techniques such as multi-scale feature refinement, local attention, and global attention to improve their performance. The specific applications of these models range from defect detection in steel and other materials to image restoration of ancient texts and rock art.