Homewear Image Classification Dataset

The current retail e-commerce industry faces challenges such as low product recognition efficiency and poor user experience. Especially in the homewear market, consumers rely on high-quality image data for an intuitive understanding of products. However, existing image classification datasets are limited and cannot meet the demand for precise classification in the market. This dataset aims to address the issues of accuracy and diversity in image classification by providing a rich collection of homewear product images. The dataset is constructed by using web crawlers to capture homewear images from various e-commerce platforms, ensuring sample diversity. Each image undergoes multiple rounds of manual annotation and consistency checks, with final reviews conducted by industry experts to ensure data quality. The data is stored in JPG format, organized with a mapping relationship between image IDs and classification labels, facilitating subsequent model training and application.