Generative Adversarial Image Super‐Resolution Through Deep Dense Skip Connections

Zhu, Xiaobin; Li, Zhuangzi; Zhang, Xiaoyu; Li, Haisheng; Xue, Ziyu; Wang, Lei

doi:10.1111/cgf.13568

Cited by 26 publications

(11 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, jagging and hallucination artifacts often occur near the strong edge. In [54], the authors propose deep dense skip connection based generator to reduce over-smoothing and use the Wasserstein distance based discriminator to improve the stability of learning. They achieve higher PSNR, SSIM and MOS with less blur.…”

Section: Related Workmentioning

confidence: 99%

“…In summary, the studies that propose a novel discriminator [21], [22], [54] have successfully reduced the blur artifact, but they have suffered from noise generation, which degrades the image quality.…”

Section: Related Workmentioning

confidence: 99%

“…And the simultaneous use of both GMD and DCT discriminators reduces both artifacts and makes the reconstructed HF closer to ground-truth. Figures 11-13 show comparisons of the proposed method with the existing methods described in [7], [9], [18], [21], [54]. In Figures 11-13, the SS-SR results [7] show the limits of non-deep-learning-based approaches.…”

Section: Visual Comparisonmentioning

confidence: 99%

See 2 more Smart Citations

Multi-Perspective Discriminators-Based Generative Adversarial Network for Image Super Resolution

2019

View full text Add to dashboard Cite

Recently, generative adversarial network-based image super resolution has been investigated, and it has been shown to lead to overwhelming improvements in subjective quality. However, it also leads to checkerboard artifacts and the unpleasing high-frequency (HF) components. In this paper, we propose a multi-discriminators-based image super resolution method that distinguishes those artifacts from various perspectives. First, the DCT perspective discriminator is proposed because the checkerboard artifacts are easily separated on the frequency domain. Second, the gradient perspective discriminator is proposed, because the unpleasing HF components can be discriminated on the gradient magnitude distribution. These proposed multi-perspective discriminators can easily identify artifacts, and they can help the generator reproduce artifact-less SR images. The experimental results show that the proposed SR-GAN with multiperspective discriminators achieves objective and subjective quality improvements in terms of PSNR, SSIM, PI and MOS, as compared to the conventional SR-GAN by reducing the aforementioned artifacts. INDEX TERMS Image super-resolution, deep learning for super resolution, SR GAN, multi-discriminators.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Visual Comparisonmentioning

confidence: 99%

See 1 more Smart Citation

Multi-Perspective Discriminators-Based Generative Adversarial Network for Image Super Resolution

2019

View full text Add to dashboard Cite

show abstract

“…Thus, the image classification results are better. In order to solve the problem of incomplete loss function and incompetent structure in existing networks Zhu et al [68] proposed a generative adversarial image super-resolution method through deep dense skip connections (GSR-DDNet). This method model the mapping across the low-quality and highquality images in an adversarial way.…”

Section: Generative Adersarial Networkmentioning

confidence: 99%

WGAN-Based Robust Occluded Facial Expression Recognition

et al. 2019

View full text Add to dashboard Cite

Research on facial expression recognition (FER) technology can promote the development of theoretical and practical applications for our daily life. Currently, most of the related works on this technology are focused on un-occluded FER. However, in real life, facial expression images often have partial occlusion; therefore, the accurate recognition of occluded facial expression images is a topic that should be explored. In this paper, we proposed a novel Wasserstein generative adversarial network-based method to perform occluded FER. After complementing the face occlusion image with complex facial expression information, the recognition is achieved by learning the facial expression features of the images. This method consists of a generator G and two discriminators D 1 and D 2. The generator naturally complements occlusion in the expression image under the triple constraints of weighted reconstruction loss l wr , triplet loss l t , and adversarial loss l a. We optimize the discriminator D 1 to distinguish between real and fake by constructing an adversarial loss l a between the generated complementing images, original un-occluded images, and smallscale-occluded images based on the Wasserstein distance. Finally, the FER is completed by introducing classification loss l c into D 2. To verify the effectiveness of the proposed method, an experimental analysis was performed on the AffectNet and RAF-DB datasets. The visual occlusion complementing results, comparison of recognition rates of facial expression images with and without de-occlusion processing, and T-distributed stochastic neighbor embedding visual analysis of facial expression features all prove the effectiveness of the proposed method. The experimental results show that the proposed method is better than the existing state-of-the-art methods. INDEX TERMS Facial expression recognition, partial occlusion, image complementation, Wasserstein generative adversarial network.

show abstract

“…In general, image classification is aimed at distinguishing the image categories according to their semantic information. It is widely applied in real-world applications, including face recognition [1], traffic sign detection [2], and other applications [3][4][5][6][7]. Traditional image classification methods often adopt hand-crafted features (e.g., SIFT [8] and HOG [9]) combined with classical classifiers (e.g., SVM [10]).…”

Section: Introductionmentioning

confidence: 99%

Label Rectification Learning through Kernel Extreme Learning Machine

Cai

Chen

et al. 2021

Wireless Communications and Mobile Computing

Self Cite

View full text Add to dashboard Cite

Along with the strong representation of the convolutional neural network (CNN), image classification tasks have achieved considerable progress. However, majority of works focus on designing complicated and redundant architectures for extracting informative features to improve classification performance. In this study, we concentrate on rectifying the incomplete outputs of CNN. To be concrete, we propose an innovative image classification method based on Label Rectification Learning (LRL) through kernel extreme learning machine (KELM). It mainly consists of two steps: (1) preclassification, extracting incomplete labels through a pretrained CNN, and (2) label rectification, rectifying the generated incomplete labels by the KELM to obtain the rectified labels. Experiments conducted on publicly available datasets demonstrate the effectiveness of our method. Notably, our method is extensible which can be easily integrated with off-the-shelf networks for improving performance.

show abstract

Generative Adversarial Image Super‐Resolution Through Deep Dense Skip Connections

Cited by 26 publications

References 25 publications

Multi-Perspective Discriminators-Based Generative Adversarial Network for Image Super Resolution

Multi-Perspective Discriminators-Based Generative Adversarial Network for Image Super Resolution

WGAN-Based Robust Occluded Facial Expression Recognition

Label Rectification Learning through Kernel Extreme Learning Machine

Contact Info

Product

Resources

About