[Download] [Results] [Codes]

Mask region segmentation is a preliminary stage to tackle the occlusion issue corresponding to the face-related tasks. Existing masked face datasets are not procedure binary segmentation maps because Segmenting mask regions manually is a time-consuming operation. As a result, existing unmasking methods by overlaying masks on existing face datasets. However, since these techniques rely on an artificially generated mask, their effects tend to seem unnatural. To address this issue, the masked face segmentation dataset (MFSD) provides the first public training dataset for the face mask segmentation task.

Details on the dataset can be found at ABANet: Attention boundary-aware network for image segmentation.
If you have questions, or any suggestions to help us improve the dataset, please contact [email protected].

          ├─ /1
          │　├─ face_crop
          │　└─ face_crop_segmentation
MSFD------│　└─ img
          |  └─ dataset.csv
          ├─ /2
          │　├─ img
          

Dataset can be downloaded at GoogleDrive.

MSFD is distributed under the MIT License

Please consider citing our work:

@article{rezvaniabanet,
  title={ABANet: Attention boundary-aware network for image segmentation},
  author={Rezvani, Sadjad and Fateh, Mansoor and Khosravi, Hossein},
  journal={Expert Systems},
  pages={e13625},
  publisher={Wiley Online Library}
}

A special thanks to Yasin Rezvani for his invaluable assistance in data collection and labeling. Yasin played a crucial role in gathering data from Google and Instagram and preparing it for deep learning model training.