Upper Eyelid Segmentation Dataset

The progress of video and photo editing applications relies heavily on high-quality datasets for machine learning models training. Our carefully selected datasets play a crucial role in enhancing the abilities of such applications, by offering accurate segmentation and recognition of different elements within images and videos.

The "Upper Eyelid Segmentation Dataset" is designed for the beauty and visual entertainment industries, incorporating internet-collected images with resolutions from 100 x 100 to 400 x 400 pixels. This focused dataset is dedicated to semantic segmentation of the upper eyelid, with annotations covering both eyes, facilitating detailed eye makeup applications and character modeling.