This paper proposes a novel strategy that employs Generative Adversarial Networks (GANs) to augment data in the image segmentation field, and a Convolutional-Deconvolutional Neural Network (CDNN) to automatically generate lesion segmentation mask from dermoscopic images. Training the CDNN with our GAN generated data effectively improves the state-of-the-art.
Improving Skin Lesion Segmentation with Generative Adversarial Networks / Pollastri, Federico; Bolelli, Federico; Paredes, Roberto; Grana, Costantino. - 2018-:(2018), pp. 442-443. (Intervento presentato al convegno 31st IEEE International Symposium on Computer-Based Medical Systems, CBMS 2018 tenutosi a Karlstad, Sweden nel Jun 18-21) [10.1109/CBMS.2018.00086].
Improving Skin Lesion Segmentation with Generative Adversarial Networks
Federico Pollastri;Federico Bolelli
;Costantino Grana
2018
Abstract
This paper proposes a novel strategy that employs Generative Adversarial Networks (GANs) to augment data in the image segmentation field, and a Convolutional-Deconvolutional Neural Network (CDNN) to automatically generate lesion segmentation mask from dermoscopic images. Training the CDNN with our GAN generated data effectively improves the state-of-the-art.
| File | Dimensione | Formato | |
|---|---|---|---|
|
CBMS_2018_paper_177.pdf
Open access
Tipologia:
Versione dell'autore revisionata e accettata per la pubblicazione
Dimensione
596.94 kB
Formato
Adobe PDF
|
596.94 kB | Adobe PDF | Visualizza/Apri |
Pubblicazioni consigliate
I metadati presenti in IRIS UNIMORE sono rilasciati con licenza Creative Commons CC0 1.0 Universal, mentre i file delle pubblicazioni sono rilasciati con licenza Attribuzione 4.0 Internazionale (CC BY 4.0), salvo diversa indicazione.
In caso di violazione di copyright, contattare Supporto Iris
