LIAGM/DeepFashion-MultiModal-Parts2Whole

DeepFashion MultiModal Parts2Whole Dataset Details Dataset Description This human image dataset comprising about 41,500 reference-target pairs. Each pair in this dataset includes multiple reference images, which encompass human pose images (e.g., OpenPose, Human Parsing, DensePose), various aspects of human appearance (e.g., hair, face, clothes, shoes) with their short textual labels, and a target image featuring the same individual (ID) in the… See the full description on the dataset page: https://huggingface.co/datasets/LIAGM/DeepFashion-MultiModal-Parts2Whole.