Household Cleaning Products Occlusion Image Dataset

The retail e-commerce industry faces significant challenges in accurately recognizing products that are partially obstructed in images. Current models often struggle with occlusion, leading to decreased performance in product detection and search functionalities. Existing datasets lack sufficient diversity in occluded images, making it difficult for machine learning models to generalize. This dataset aims to address these challenges by providing a rich collection of images featuring common household cleaning products like Lysol, Clorox, Mr. Clean, and Vanish, captured in various occlusion scenarios. The data was collected using high-resolution cameras in controlled environments to ensure clarity and consistency. Quality control measures included multiple rounds of annotation, consistency checks among annotators, and expert reviews to ensure high accuracy. The dataset is organized in JPG format, with images structured in folders by product type, facilitating easy access and integration into machine learning workflows. The dataset contains 15,000 images, and the file size is estimated at 1.8 GB.