TY - GEN
T1 - ABCAS
T2 - 2023 IEEE International Conference on Consumer Electronics, ICCE 2023
AU - Hirose, Shota
AU - Maki, Shiori
AU - Wada, Naoki
AU - Katto, Jiro
AU - Sun, Heming
N1 - Publisher Copyright:
© 2023 IEEE.
PY - 2023
Y1 - 2023
N2 - Spectral Normalization is one of the best methods for stabilizing the training of Generative Adversarial Network. Spectral Normalization limits the gradient of discriminator between the distribution between real data and fake data. However, even with this normalization, GAN's training sometimes fails. In this paper, we reveal that more severe restriction is sometimes needed depending on the training dataset, then we propose a novel stabilizer which offers an adaptive normalization method, called ABCAS. Our method decides discriminator's Lipschitz constant adaptively, by checking the distance of distributions of real and fake data. Our method improves the stability of the training of Generative Adversarial Network and achieved better Fréchet Inception Distance score of generated images. We also investigated suitable spectral norm for three datasets. We show the result as an ablation study.
AB - Spectral Normalization is one of the best methods for stabilizing the training of Generative Adversarial Network. Spectral Normalization limits the gradient of discriminator between the distribution between real data and fake data. However, even with this normalization, GAN's training sometimes fails. In this paper, we reveal that more severe restriction is sometimes needed depending on the training dataset, then we propose a novel stabilizer which offers an adaptive normalization method, called ABCAS. Our method decides discriminator's Lipschitz constant adaptively, by checking the distance of distributions of real and fake data. Our method improves the stability of the training of Generative Adversarial Network and achieved better Fréchet Inception Distance score of generated images. We also investigated suitable spectral norm for three datasets. We show the result as an ablation study.
KW - Adaptive Scheduling
KW - Generative Adversarial Network
KW - Lipschitz constant
KW - Spectral Normalization
UR - http://www.scopus.com/inward/record.url?scp=85149152838&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85149152838&partnerID=8YFLogxK
U2 - 10.1109/ICCE56470.2023.10043368
DO - 10.1109/ICCE56470.2023.10043368
M3 - Conference contribution
AN - SCOPUS:85149152838
T3 - Digest of Technical Papers - IEEE International Conference on Consumer Electronics
BT - 2023 IEEE International Conference on Consumer Electronics, ICCE 2023
PB - Institute of Electrical and Electronics Engineers Inc.
Y2 - 6 January 2023 through 8 January 2023
ER -