Casual Jacket Occlusion Image Dataset

The retail e-commerce sector faces challenges such as accurately detecting clothing items when they are partially obscured, leading to potential misclassification and poor customer experiences. Existing datasets often lack diversity in occlusion scenarios and do not adequately represent real-world conditions where clothing items are worn. This dataset aims to address these issues by providing a comprehensive collection of images where casual jackets are occluded by movement or other clothing items. The data was collected using high-resolution cameras in varied lighting conditions, ensuring a wide range of scenarios. Quality control measures included multiple rounds of annotation by trained professionals, followed by consistency checks and expert reviews. The dataset is organized in JPG format, with images systematically categorized for ease of access and analysis.