Progressive Growing of GANs for Improved Quality, Stability, and Variation

Karras, Tero; Aila, Timo; Laine, Samuli; Lehtinen, Jaakko

doi:10.48550/arxiv.1710.10196

Cited by 1,167 publications

(1,852 citation statements)

References 11 publications

Supporting

Mentioning

1,841

Contrasting

Unclassified

Order By: Relevance

“…StyleGAN [24] combines the architecture of Progressive Growing GAN [25] with the style transfer principles [26] in Figure 2. StyleGAN's architecture addresses some limitations of the GAN models, such as stability during training and lack of control over the images generated.…”

Section: Theoretical Background 21 Styleganmentioning

confidence: 99%

StyleGANs and Transfer Learning for Generating Synthetic Images in Industrial Applications

2021

View full text Add to dashboard Cite

Deep learning applications on computer vision involve the use of large-volume and representative data to obtain state-of-the-art results due to the massive number of parameters to optimise in deep models. However, data are limited with asymmetric distributions in industrial applications due to rare cases, legal restrictions, and high image-acquisition costs. Data augmentation based on deep learning generative adversarial networks, such as StyleGAN, has arisen as a way to create training data with symmetric distributions that may improve the generalisation capability of built models. StyleGAN generates highly realistic images in a variety of domains as a data augmentation strategy but requires a large amount of data to build image generators. Thus, transfer learning in conjunction with generative models are used to build models with small datasets. However, there are no reports on the impact of pre-trained generative models, using transfer learning. In this paper, we evaluate a StyleGAN generative model with transfer learning on different application domains—training with paintings, portraits, Pokémon, bedrooms, and cats—to generate target images with different levels of content variability: bean seeds (low variability), faces of subjects between 5 and 19 years old (medium variability), and charcoal (high variability). We used the first version of StyleGAN due to the large number of publicly available pre-trained models. The Fréchet Inception Distance was used for evaluating the quality of synthetic images. We found that StyleGAN with transfer learning produced good quality images, being an alternative for generating realistic synthetic images in the evaluated domains.

show abstract

Section: Theoretical Background 21 Styleganmentioning

confidence: 99%

StyleGANs and Transfer Learning for Generating Synthetic Images in Industrial Applications

2021

View full text Add to dashboard Cite

show abstract

“…CelebAMask-HQ [17] comes closest to meeting these requirements among the many face segmentation datasets. It provides manually annotated masks over 30K high-resolution face images in CelebA-HQ [13], covering 18 facial attributes. Nevertheless, this dataset was designed for a GAN-based face image editing task, thus ignoring the various occlusions on the face image (except for hair and glasses), e.g., it wrongly annotates occlusions such as hands as attributes that they obscure.…”

Section: Introductionmentioning

confidence: 99%

FaceOcc: A Diverse, High-quality Face Occlusion Dataset for Human Face Extraction

Yin,

Chen

2022

Preprint

View full text Add to dashboard Cite

Occlusions often occur in face images in the wild, troubling face-related tasks such as landmark detection, 3D reconstruction, and face recognition. It is beneficial to extract face regions from unconstrained face images accurately. However, current face segmentation datasets suffer from small data volumes, few occlusion types, low resolution, and imprecise annotation, limiting the performance of data-driven-based algorithms. This paper proposes a novel face occlusion dataset with manually labeled face occlusions from the CelebA-HQ and the internet. The occlusion types cover sunglasses, spectacles, hands, masks, scarfs, microphones, etc. To the best of our knowledge, it is by far the largest and most comprehensive face occlusion dataset. Combining it with the attribute mask in CelebAMask-HQ, we trained a straightforward face segmentation model but obtained SOTA performance, convincingly demonstrating the effectiveness of the proposed dataset.

show abstract

“…Since a goal of GAN is discovering the Nash equilibrium of non-convex game in the high-dimensional parameter space, GAN is substantially more complex and difficult to train compared to networks trained by supervised learning [49]. To address this issue, some papers [4,15,48] investigated novel network architectures for discriminator and generator. Although these methods can produce high-resolution images on challenging datasets such as ImageNet [20], they still have the fundamental problem related to the training instability.…”

Section: Introductionmentioning

confidence: 99%

Image Generation with Self Pixel-wise Normalization

Yeo¹,

Sagong²,

Park³

et al. 2022

Preprint

View full text Add to dashboard Cite

Region-adaptive normalization (RAN) methods have been widely used in the generative adversarial network (GAN)-based image-to-image translation technique. However, since these approaches need a mask image to infer the pixel-wise affine transformation parameters, they cannot be applied to the general image generation models having no paired mask images. To resolve this problem, this paper presents a novel normalization method, called self pixel-wise normalization (SPN), which effectively boosts the generative performance by performing the pixel-adaptive affine transformation without the mask image. In our method, the transforming parameters are derived from a self-latent mask that divides the feature map into the foreground and background regions. The visualization of the selflatent masks shows that SPN effectively captures a sin-

show abstract

Progressive Growing of GANs for Improved Quality, Stability, and Variation

Cited by 1,167 publications

References 11 publications

StyleGANs and Transfer Learning for Generating Synthetic Images in Industrial Applications

StyleGANs and Transfer Learning for Generating Synthetic Images in Industrial Applications

FaceOcc: A Diverse, High-quality Face Occlusion Dataset for Human Face Extraction

Image Generation with Self Pixel-wise Normalization

Contact Info

Product

Resources

About