This dataset has been generated using images from other datasets and others obtained from trusted webpages (like CDC), and trimmed to square skin sections of 224x224 pixels.
The images contained in this dataset will continue to increase with information from other datasets and/or trusted webpages.
Authors responsible for creating this dataset belong to "Architecture and Computer Technology" department from Universidad de Sevilla (Spain).
In order to download this dataset, you only have to fill the MonkeypoxSkin EULA document attached to this webpage and send it to the email address indicated in that document.
The dataset you can download is formed by three classes: MONKEYPOX, NORMAL (HEALTHY) and OTHER SKIN DAMAGES. Each class is formed by 100 images (and increasing) of close superficial skin tissue.
Here you can see some examples from each class:
| Monkeypox | Normal | Others |
|---|---|---|
 |
 |
 |
 |
 |
 |
 |
 |
 |
The work were this dataset is described is "Monkeypox Diagnostic-Aid System with Skin Images Using Convolutional Neural Networks" and can be accesed in pre-print mode using this link: http://ssrn.com/abstract=4186534
You can cite this work using this suggested citation:
Muñoz-Saavedra, Luis and Escobar-Linero, Elena and Civit-Masot, Javier and Luna-Perejón, Francisco and Civit, Antón and Domínguez-Morales, Manuel, Monkeypox Diagnostic-Aid System with Skin Images Using Convolutional Neural Networks. Available at SSRN: https://ssrn.com/abstract=4186534 or http://dx.doi.org/10.2139/ssrn.4186534
For suggestions and questions you can contact Prof. Manuel Domínguez-Morales (mjdominguez@us.es).