Filters








249 Hits in 4.8 sec

EDPN: Enhanced Deep Pyramid Network for Blurry Image Restoration [article]

Ruikang Xu, Zeyu Xiao, Jie Huang, Yueyi Zhang, Zhiwei Xiong
2021 arXiv   pre-print
Experimental results demonstrate that our method significantly outperforms existing solutions for blurry image super-resolution and blurry image deblocking.  ...  Image deblurring has seen a great improvement with the development of deep neural networks.  ...  [63] train a generative adversarial network to super-resolve blurry face and text images. Zhang et al.  ... 
arXiv:2105.04872v1 fatcat:4jjp7s5pjbd5flw7ar3sed46qu

Face hallucination using cascaded super-resolution and identity priors

Klemen Grm, Walter J. Scheirer, Vitomir Struc
2019 IEEE Transactions on Image Processing  
The model consists of two main parts: i) a cascaded super-resolution network that upscales the low-resolution facial images, and ii) an ensemble of face recognition models that act as identity priors for  ...  Different from most competing super-resolution techniques that rely on a single model for upscaling (even with large magnification factors), our network uses a cascade of multiple SR models that progressively  ...  The Cascaded SR Network The generative part of the C-SRIP model, the cascaded SR network, is a 52-layer CNN that takes a LR facial image as input and super-resolves it at a magnification factor of 8×.  ... 
doi:10.1109/tip.2019.2945835 pmid:31613762 fatcat:35qwxawe4ba4bajtbum2jvgu7q

Joint Face Hallucination and Deblurring via Structure Generation and Detail Enhancement [article]

Yibing Song, Jiawei Zhang, Lijun Gong, Shengfeng He, Linchao Bao, Jinshan Pan, Qingxiong Yang, Ming-Hsuan Yang
2018 arXiv   pre-print
We address the problem of restoring a high-resolution face image from a blurry low-resolution input. This problem is difficult as super-resolution and deblurring need to be tackled simultaneously.  ...  We first propose a facial component guided deep Convolutional Neural Network (CNN) to restore a coarse face image, which is denoted as the base image where the facial component is automatically generated  ...  A cascaded bi-network is proposed in [66] for FH and deep reinforcement learning is applied in [2] to achieve attention awareness.  ... 
arXiv:1811.09019v1 fatcat:zhaokqukyjfl5ocamyz67cbd4i

Self-Enhanced Convolutional Network for Facial Video Hallucination [article]

Chaowei Fang, Guanbin Li, Xiaoguang Han, Yizhou Yu
2019 arXiv   pre-print
As a domain-specific super-resolution problem, facial image hallucination has enjoyed a series of breakthroughs thanks to the advances of deep convolutional neural networks.  ...  Taking advantage of high inter-frame dependency in videos, we propose a self-enhanced convolutional network for facial video hallucination.  ...  [21] super-resolves every LR image using multiple frames via learning a dynamic upsampling filter for each pixel in the target HR image and a residual image.  ... 
arXiv:1911.11136v1 fatcat:xavmtdckufa6ngargrf4lspv5a

Efficient Deep Neural Network for Photo-realistic Image Super-Resolution [article]

Namhyuk Ahn, Byungkon Kang, Kyung-Ah Sohn
2022 arXiv   pre-print
In detail, we design an architecture that implements a cascading mechanism on a residual network to boost the performance with limited resources via multi-level feature fusion.  ...  Recent progress in deep learning-based models has improved photo-realistic (or perceptual) single-image super-resolution significantly.  ...  Table 1 shows the model analysis on the effect of cascading modules and the global residual learning scheme. Model with local cascading improves the baseline SR performance.  ... 
arXiv:1903.02240v5 fatcat:kb4buwdqbbbwten2ojxczqpcgy

URNet: A U-Shaped Residual Network for Lightweight Image Super-Resolution

Yuntao Wang, Lin Zhao, Liman Liu, Huaifei Hu, Wenbing Tao
2021 Remote Sensing  
To address this problem, we introduce a lightweight U-shaped residual network (URNet) for fast and accurate image SR.  ...  Moreover, a lightweight asymmetric residual non-local block is proposed to model the global context information and further improve the performance of SR.  ...  Acknowledgments: The authors are grateful to the Editor and reviewers for their constructive comments, which significantly improved this work.  ... 
doi:10.3390/rs13193848 fatcat:odyf56rg45grvnm4a52fv3adtm

Face Recognition in Low Quality Images: A Survey [article]

Pei Li, Loreto Prieto, Domingo Mery, Patrick Flynn
2019 arXiv   pre-print
Several studies addressed this problem employed techniques like super resolution, deblurring, or learning a relationship between different resolution domains.  ...  Finally, we summarized the general limitations and speculate a priorities for the future effort.  ...  Dan et.al [83] designed a model using deep architecture and mixed LR images with HR images for training which achieved learning a highly non-linear resolution invariant space.  ... 
arXiv:1805.11519v3 fatcat:izpl554u3fga5d62e6jxw4zuwu

Attention-Aware Linear Depthwise Convolution for Single Image Super-Resolution [article]

Seongmin Hwang, Gwanghuyn Yu, Cheolkon Jung, Jinyoung Kim
2019 arXiv   pre-print
Specifically, linear depthwise convolution allows CNN-based SR models to preserve useful information for reconstructing a super-resolved image while reducing computational burden.  ...  In this paper, we propose an attention-aware linear depthwise network to address the problems for single image SR, named ALDNet.  ...  As shown in the figure, two types of residual learning are used to construct the network: (1) global residual learning [3] to provide skip-connection in global scale, and (2) local residual learning  ... 
arXiv:1908.02648v3 fatcat:id5vw5rmwjgdxocv32ifpmb26a

NTIRE 2020 Challenge on Image and Video Deblurring

Seungjun Nah, Sanghyun Son, Radu Timofte, Kyoung Mu Lee, Yu Tseng, Yu-Syuan Xu, Cheng-Ming Chiang, Yi-Min Tsai, Stephan Brehm, Sebastian Scherer, Dejia Xu, Yihao Chu (+36 others)
2020 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)  
On Track 2, the image deblurring methods are executed on a mobile platform to find the balance of the running speed and the restoration accuracy.  ...  Track 1 aims to develop single-image deblurring methods focusing on restoration quality.  ...  Shortcuts and residual connections at different scale effectively resolve the vanishing gradients and help the network learn more key features.  ... 
doi:10.1109/cvprw50498.2020.00216 dblp:conf/cvpr/NahSTLTXCTBSXCS20 fatcat:a6ojyfuidrbb3avwdpv4mje77e

High-Order Residual Network for Light Field Super-Resolution

Nan Meng, Xiaofei Wu, Jianzhuang Liu, Edmund Lam
2020 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
In this paper, we propose a novel high-order residual network to learn the geometric features hierarchically from the LF for reconstruction.  ...  After fully obtaining the local features learned from each HRB, our model extracts the representative geometric features for spatio-angular upsampling through the global residual learning.  ...  Geometric representation learning network The GRLNet is composed of a set of cascaded HRBs.  ... 
doi:10.1609/aaai.v34i07.6847 fatcat:44yhsw3f4zb55exfjy5lc6ib6m

Deep Laplacian Pyramid Networks for Fast and Accurate Super-Resolution [article]

Wei-Sheng Lai, Jia-Bin Huang, Narendra Ahuja, Ming-Hsuan Yang
2017 arXiv   pre-print
Convolutional neural networks have recently demonstrated high-quality reconstruction for single-image super-resolution.  ...  In this paper, we propose the Laplacian Pyramid Super-Resolution Network (LapSRN) to progressively reconstruct the sub-band residuals of high-resolution images.  ...  [7] propose a Super-Resolution Convolutional Neural Network (SRCNN) to learn a nonlinear LR-to-HR mapping.  ... 
arXiv:1704.03915v2 fatcat:n6lhowwtfzcipol5jy3logsgrq

Deep Laplacian Pyramid Networks for Fast and Accurate Super-Resolution

Wei-Sheng Lai, Jia-Bin Huang, Narendra Ahuja, Ming-Hsuan Yang
2017 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
In this paper, we propose the Laplacian Pyramid Super-Resolution Network (LapSRN) to progressively reconstruct the sub-band residuals of high-resolution images.  ...  Convolutional neural networks have recently demonstrated high-quality reconstruction for single-image superresolution.  ...  [7] propose a Super-Resolution Convolutional Neural Network (SRCNN) to learn a nonlinear LR-to-HR mapping.  ... 
doi:10.1109/cvpr.2017.618 dblp:conf/cvpr/LaiHA017 fatcat:gviav6i45rddjp6ucysk3zqwcq

Channel-wise and Spatial Feature Modulation Network for Single Image Super-Resolution [article]

Yanting Hu, Jie Li, Yuanfei Huang, Xinbo Gao
2018 arXiv   pre-print
To capture more informative features and maintain long-term information for image super-resolution, we propose a channel-wise and spatial feature modulation (CSFM) network in which a sequence of feature-modulation  ...  On the other hand, as the depth of neural networks grows, the long-term information coming from preceding layers is easy to be weaken or lost in late layers, which is adverse to super-resolving image.  ...  A. Deep-learning based Image Super-Resolution Since Dong et al.  ... 
arXiv:1809.11130v1 fatcat:r6f2nb5knvayjayzz33tboapqe

Deep Residual Networks with a Fully Connected Recon-struction Layer for Single Image Super-Resolution [article]

Yongliang Tang, Jiashui Huang, Faen Zhang, Weiguo Gong
2019 arXiv   pre-print
Recently, deep neural networks have achieved impressive performance in terms of both reconstruction accuracy and efficiency for single image super-resolution (SISR).  ...  However, the network model of these methods is a fully convolutional neural network, which is limit to exploit the differentiated contextual information over the global region of the input image because  ...  We resolve this issue by incorporating the convolution layers into the cascaded residual units and constructing local paths for propagating information directly.  ... 
arXiv:1805.10143v2 fatcat:dcyoxpw5nvehri6qtthzto2f3y

NTIRE 2020 Challenge on Image and Video Deblurring [article]

Seungjun Nah, Sanghyun Son, Radu Timofte, Kyoung Mu Lee
2020 arXiv   pre-print
On Track 2, the image deblurring methods are executed on a mobile platform to find the balance of the running speed and the restoration accuracy.  ...  Track 1 aims to develop single-image deblurring methods focusing on restoration quality.  ...  Figure 4 . 4 Shortcuts and residual connections at different scale effectively resolve the vanishing gradients and help the network learn more key features.  ... 
arXiv:2005.01244v2 fatcat:aoy3tyxlybefrd7yd5ywvr6jh4
« Previous Showing results 1 — 15 out of 249 results