Jacket Occlusion Image Dataset

The retail e-commerce industry faces significant challenges in accurately recognizing and classifying apparel items when they are occluded by jackets or other outerwear. Existing solutions often struggle with occlusion detection, leading to misclassification and poor user experience. This dataset aims to address specific technical issues related to occlusion in images, enabling better object detection algorithms and improving the accuracy of automated systems in recognizing partially hidden clothing items. The dataset is collected using high-resolution cameras in various retail environments, capturing images with diverse lighting and angles. A rigorous quality control process is employed, which includes multiple rounds of annotations, consistency checks among different annotators, and expert reviews to ensure high annotation precision. The images are stored in JPG format and organized in a structured directory based on occlusion types and categories.