Generative Adversarial Networks (GAN) are being increasingly used to perform face aging due to their capabilities of automatically generating highly-realistic synthetic images by using an adversarial model often based on Convolutional Neural Networks (CNN). However, GANs currently represent black box models since it is not known how the CNNs store and process the information learned from data. In this paper, we propose the first method that deals with explaining GANs, by introducing a novel qualitative and quantitative analysis of the inner structure of the model. Similarly to analyzing the common genes in two DNA sequences, we analyze the common filters in two CNNs. We show that the GANs for face aging partially share their parameters with GANs trained for heterogeneous applications and that the aging transformation can be learned using general purpose image databases and a fine-tuning step. Results on public databases confirm the validity of our approach, also enabling future studies on similar models.
Towards explainable face aging with Generative Adversarial Networks / A. Genovese, V. Piuri, F. Scotti (PROCEEDINGS - INTERNATIONAL CONFERENCE ON IMAGE PROCESSING). - In: 2019 IEEE International Conference on Image Processing (ICIP)[s.l] : IEEE, 2019. - ISBN 9781538662496. - pp. 3806-3810 (( Intervento presentato al 26. convegno International Conference on Image Processing tenutosi a Taipei nel 2019.
Towards explainable face aging with Generative Adversarial Networks
A. Genovese;V. Piuri;F. Scotti
2019
Abstract
Generative Adversarial Networks (GAN) are being increasingly used to perform face aging due to their capabilities of automatically generating highly-realistic synthetic images by using an adversarial model often based on Convolutional Neural Networks (CNN). However, GANs currently represent black box models since it is not known how the CNNs store and process the information learned from data. In this paper, we propose the first method that deals with explaining GANs, by introducing a novel qualitative and quantitative analysis of the inner structure of the model. Similarly to analyzing the common genes in two DNA sequences, we analyze the common filters in two CNNs. We show that the GANs for face aging partially share their parameters with GANs trained for heterogeneous applications and that the aging transformation can be learned using general purpose image databases and a fine-tuning step. Results on public databases confirm the validity of our approach, also enabling future studies on similar models.File | Dimensione | Formato | |
---|---|---|---|
icip19.pdf
accesso aperto
Tipologia:
Post-print, accepted manuscript ecc. (versione accettata dall'editore)
Dimensione
1.1 MB
Formato
Adobe PDF
|
1.1 MB | Adobe PDF | Visualizza/Apri |
08803616.pdf
accesso riservato
Tipologia:
Publisher's version/PDF
Dimensione
1.15 MB
Formato
Adobe PDF
|
1.15 MB | Adobe PDF | Visualizza/Apri Richiedi una copia |
Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.