PConv: simple yet effective convolutional layer for generative adversarial network

2021/01/31 by Seung Park, Yoon-Jae Yeo, Yong-Goo Shin · 1 citation
Computer Science · Mathematics · #Advanced Image Processing Techniques #Algorithm #Artificial intelligence #Artificial neural network #Computer science #Convolution (computer science) #Convolutional neural network #Discriminator #Generalization #Generative Adversarial Networks and Image Synthesis #Human Pose and Action Recognition #Layer (electronics) #Lipschitz continuity #Mathematical analysis #Mathematics #Memorization #Pure mathematics #Simple (philosophy) #Tensor (intrinsic definition) #cs.AI #cs.CV #cs.LG

paper · pdf · doi:10.1007/s00521-021-06846-2

published in Neural Computing and Applications 34(9), 7113-7124 (Springer Science+Business Media) · Submitted to journal, arXiv admin note: text overlap with arXiv:1911.10979

arxiv created 2021/11/07 · arxiv updated 2021/12/23 · openalex publication_date 2022/01/27 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/05

Abstract

This paper presents a novel convolutional layer, called perturbed convolution (PConv), which focuses on achieving two goals simultaneously: improving the generative adversarial network (GAN) performance and alleviating the memorization problem in which the discriminator memorizes all images from a given dataset as training progresses. In PConv, perturbed features are generated by randomly disturbing an input tensor before performing the convolution operation. This approach is simple but surprisingly effective. First, to produce a similar output even with the perturbed tensor, each layer in the discriminator should learn robust features having a small local Lipschitz value. Second, since the input tensor is randomly perturbed during the training procedure like the dropout in neural networks, the memorization problem could be alleviated. To show the generalization ability of the proposed method, we conducted extensive experiments with various loss functions and datasets including CIFAR-10, CelebA, CelebA-HQ, LSUN, and tiny-ImageNet. The quantitative evaluations demonstrate that PConv effectively boosts the performance of GAN and conditional GAN in terms of Frechet inception distance (FID).

Citations

• Simple yet Effective Way for Improving the Performance of GAN
• LSUN: Construction of a Large-scale Image Dataset using Deep Learning with Humans in the Loop
• Adam: A Method for Stochastic Optimization
• Conditional Generative Adversarial Nets
• A Learned Representation For Artistic Style
• Conditional Image Synthesis With Auxiliary Classifier GANs
• Improved Regularization of Convolutional Neural Networks with Cutout
• mixup: Beyond Empirical Risk Minimization
• On Convergence and Stability of GANs
• Spectral Normalization for Generative Adversarial Networks
• cGANs with Projection Discriminator
• Large Scale GAN Training for High Fidelity Natural Image Synthesis
• Consistency Regularization for Generative Adversarial Networks
• Image Augmentations for GAN Training
• On Data Augmentation for GAN Training

Cited by

• The Tenth NTIRE 2025 Efficient Super-Resolution Challenge Report