Cata7 is the first cataract surgical instrument dataset for semantic segmentation. The dataset consists of seven videos, each of which fully records a cataract surgery. All videos are from Beijing Tongren Hospital. Each video is divided into a series of images with a resolution of 1920×1080 pixels. To reduce redundancy, the videos are downsampled from 30 fps to 1 fps. In addition, images without surgical instruments are manually deleted.

Each image is annotated with precise surgical instrument edges and types. The dataset contains 2,500 images, divided into training and test sets. The training set consists of five video sequences, and the test set consists of two video sequences.