Photorealistic Style Transfer via Wavelet Transforms

Yoo, Jaejun; Uh, Youngjung; Chun, Sanghyuk; Kang, Byeongkyu; Ha, Jung-Woo

doi:10.1109/iccv.2019.00913

Cited by 332 publications

(259 citation statements)

References 33 publications

Supporting

Mentioning

239

Contrasting

Order By: Relevance

“…To overcome this hurdle, Li et al 79 used wavelet transformation as well as multilevel stylization. Following this research, Yoo et al 80 devised the wavelet pooling layer to enable photorealistic style transfer.…”

Section: Image To Image Translation Without Using Generative Adversarmentioning

confidence: 99%

Deep Learning in Medical Imaging

et al. 2019

View full text Add to dashboard Cite

The artificial neural network (ANN), one of the machine learning (ML) algorithms, inspired by the human brain system, was developed by connecting layers with artificial neurons. However, due to the low computing power and insufficient learnable data, ANN has suffered from overfitting and vanishing gradient problems for training deep networks. The advancement of computing power with graphics processing units and the availability of large data acquisition, deep neural network outperforms human or other ML capabilities in computer vision and speech recognition tasks. These potentials are recently applied to healthcare problems, including computer-aided detection/diagnosis, disease prediction, image segmentation, image generation, etc. In this review article, we will explain the history, development, and applications in medical imaging

show abstract

Section: Image To Image Translation Without Using Generative Adversarmentioning

confidence: 99%

Deep Learning in Medical Imaging

et al. 2019

View full text Add to dashboard Cite

show abstract

“…Therefore, most of evaluations of neural style transfer models are qualitative [ 48 , 49 ]. The most common method is to qualitatively compare results of style transfer methods by putting stylized images side by side [ 11 , 12 , 13 , 14 , 21 , 24 , 29 ]. Besides showing stylized images, user study is also used for evaluation [ 9 , 14 , 29 ].…”

Section: Resultsmentioning

confidence: 99%

“…For example, Gatys et al [ 8 ] use the convolutional neural network (CNN) to reconstruct the content and style, and optimize the stylized image iteratively based on a loss function. From then on, CNN-based neural style transfer has become a hot topic [ 9 , 10 , 11 , 13 , 14 ].…”

Section: Introductionmentioning

confidence: 99%

Robust Nonparametric Distribution Transfer with Exposure Correction for Image Neural Style Transfer

Liu

Hong

et al. 2020

Sensors

View full text Add to dashboard Cite

Image neural style transfer is a process of utilizing convolutional neural networks to render a content image based on a style image. The algorithm can compute a stylized image with original content from the given content image but a new style from the given style image. Style transfer has become a hot topic both in academic literature and industrial applications. The stylized results of current existing models are not ideal because of the color difference between two input images and the inconspicuous details of content image. To solve the problems, we propose two style transfer models based on robust nonparametric distribution transfer. The first model converts the color probability density function of the content image into that of the style image before style transfer. When the color dynamic range of the content image is smaller than that of style image, this model renders more reasonable spatial structure than the existing models. Then, an adaptive detail-enhanced exposure correction algorithm is proposed for underexposed images. Based this, the second model is proposed for the style transfer of underexposed content images. It can further improve the stylized results of underexposed images. Compared with popular methods, the proposed methods achieve the satisfactory qualitative and quantitative results.

show abstract

“…In this section, we focus on the wavelet application of image synthesis, which is one of the most attractive issues in the community. Yoo et al (2019) propose a wavelet corrected transfer based on whitening and coloring transforms (WCT2) which encourages the transfer network to generate visually realistic images. Provenzi et al (2014) design a wavelet-based perceptually inspired color correction variational model to deal with both the color enhancement and color cast removal.…”

Section: Wavelet Domain Image Synthesismentioning

confidence: 99%

Selective Wavelet Attention Learning for Single Image Deraining

Huang

Chai³

et al. 2021

Int J Comput Vis

View full text Add to dashboard Cite

Single image deraining refers to the process of restoring the clean background scene from a rainy image. Current approaches have resorted to deep learning techniques to remove rain from a single image by leveraging some prior information. However, due to the various appearances of rain streaks and accumulation, it is difficult to separate rain and background information in the embedding space, which results in inaccurate deraining. To address this issue, this paper proposes a selective wavelet attention learning method by learning a series of wavelet attention maps to guide the separation of rain and background information in both spatial and frequency domains. The key aspect of our method is utilizing wavelet transform to learn the content and structure of rainy features because the high-frequency features are more sensitive to rain degradations, whereas the low-frequency features preserve more of the background content. To begin with, we develop a selective wavelet attention encoder-decoder network to learn wavelet attention maps guiding the separation of rainy and background features at multiple scales. Meanwhile, we introduce wavelet pooling and unpooling to the encoder-decoder network, which shows superiority in learning increasingly abstract representations while preserving the background details. In addition, we propose latent alignment learning to supervise the background features as well as augment the training data to further improve the accuracy of deraining. Finally, we employ a hierarchical discriminator network based on selective wavelet attention to adversarially improve the visual fidelity of the generated results both globally and locally. Extensive experiments on synthetic and real datasets demonstrate that the proposed approach achieves more appealing results both quantitatively and qualitatively than the recent state-of-the-art methods. Keywords Deraining • Wavelet transform • Selective attention • Wavelet pooling and unpooling Communicated by Dengxin Dai.

show abstract

Photorealistic Style Transfer via Wavelet Transforms

Cited by 332 publications

References 33 publications

Deep Learning in Medical Imaging

Deep Learning in Medical Imaging

Robust Nonparametric Distribution Transfer with Exposure Correction for Image Neural Style Transfer

Selective Wavelet Attention Learning for Single Image Deraining

Contact Info

Product

Resources

About