A Masked and Unmasked Face Dataset

Description

A Masked and Unmasked Face Dataset is a collection of a face without medical mask (unmasked face) images and a face with medical mask (masked face) images. Based on the Deep Learning approach, this dataset can be used to build a model to predict the original face (unmasked face) from a masked face image. The dataset is organized into is in the Cropped folder. The folder contains images focusing only on the face area (both masked and unmasked); the size of the images varies depending on the size of the detected face area. Please cite the paper about face image reconstruction (using this dataset) from the authors . Each main folder consists of two subfolders, i.e., Masked and Unmasked subfolders. The Masked subfolder consists of face images with a medical mask; on the other hand, the Unmasked subfolder consists of face images without a medical mask. There are 330 images in each subfolder.

Institutions

Categories

Funding

Licence