Overlapped 100 10

The Overlapped 100-10 is a dataset in the field of computer vision, designed to promote the robustness and accuracy of object detection and recognition algorithms by providing a large number of overlapping image pairs. The dataset contains 100 categories, each with 10 highly overlapping image samples, which helps in evaluating and improving model performance in complex scenarios, enhancing reliability and effectiveness in practical applications.