RuFashion-2M is a comprehensive Russian-language fashion dataset comprising nearly 2 million (1,919,733) annotated images aggregated from 10 fashion datasets (check the datasets section). This dataset is specifically designed to research in computer vision and natural language processing tasks within the fashion domain, with a unique focus on Russian-language annotations.
The dataset features rich, multi-layered annotations for each image, including:
