Filters








3,791 Hits in 4.0 sec

Deep Burst Super-Resolution [article]

Goutam Bhat and Martin Danelljan and Luc Van Gool and Radu Timofte
2021 arXiv   pre-print
We propose a novel architecture for the burst super-resolution task. Our network takes multiple noisy RAW images as input, and generates a denoised, super-resolved RGB image as output.  ...  This key advantage, along with the increasing popularity of burst photography, have made MFSR an important problem for real-world applications.  ...  Burst Super-Resolution Network In this section, we describe our burst super-resolution network. Our network inputs multiple noisy, RAW, lowresolution (LR) images captured in a single burst.  ... 
arXiv:2101.10997v2 fatcat:luxjrwxhpbcvpmxllapo5vinmq

Deep feature fusion for self-supervised monocular depth prediction [article]

Vinay Kaushik, Brejesh Lall
2020 arXiv   pre-print
Our fusion network selects features from both upper and lower levels at every level in the encoder network, thereby creating multiple feature pyramid sub-networks that are fed to the decoder after applying  ...  We also propose a refinement module learning higher scale residual depth from a combination of higher level deep features and lower level residual depth using a pixel shuffling framework that super-resolves  ...  We combine encoded features with the CoordConv solution thereby learning robust invariant features refined by a residual decoder that incorporates depth super resolution for learning fine-grained depth  ... 
arXiv:2005.07922v1 fatcat:lj7x2lteezanhfvcilz7wfxhwe

Joint Spatial and Angular Super-Resolution from a Single Image [article]

Andre Ivan, Williem, In Kyu Park
2019 arXiv   pre-print
The conventional method relies on physical-based rendering and a secondary network to solve the angular super-resolution problem.  ...  In this paper, we show that both super-resolution problems can be solved jointly from a single image by proposing a single end-to-end deep neural network that does not require a physical-based approach  ...  The idea is to jointly synthesize light field through angular and spatial super-resolution (SR) from only a single image which is abundantly available in the real world.  ... 
arXiv:1911.11619v2 fatcat:c6dvyo7bufg2xg6ip4phjx5e2y

High-Resolution Pelvic MRI Reconstruction Using a Generative Adversarial Network with Attention and Cyclic Loss [article]

Guangyuan Li, Jun Lv, Xiangrong Tong, Chengyan Wang, Guang Yang
2021 arXiv   pre-print
Therefore, we proposed a novel super-resolution method that uses a generative adversarial network (GAN) with cyclic loss and attention mechanism to generate high-resolution MR images from low-resolution  ...  Recently, super-resolution methods have shown excellent performance in accelerating MRI. In some circumstances, it is difficult to obtain high-resolution images even with prolonged scan time.  ...  The RED \cite{mao2016image} used an encoder-decoder framework in the super-resolution method.  ... 
arXiv:2107.09989v1 fatcat:4dmywsdypvbyhcefgu2lfqbx5m

Perceptually-inspired super-resolution of compressed videos [article]

Di Ma, Mariana Afonso, Fan Zhang, David R. Bull
2021 arXiv   pre-print
This approach encodes a lower resolution version of the input video and reconstructs the original resolution during decoding.  ...  In this paper, a perceptually-inspired super-resolution approach (M-SRGAN) is proposed for spatial up-sampling of compressed video using a modified CNN model, which has been trained using a generative  ...  Input Video Spatial Down- sampling Bitstream Host Encoder CNN-based Super Resolution Host Decoder Final Reconstructed Video (Full Resolution) Decoded Low Resolution Video Low Resolution  ... 
arXiv:2106.08147v1 fatcat:etpxv745dfc3rngpehzkz7iwhi

High-Resolution Pelvic MRI Reconstruction Using a Generative Adversarial Network with Attention and Cyclic Loss

Guangyuan Li, Jun Lv, Xiangrong Tong, Chengyan Wang, Guang Yang
2021 IEEE Access  
INDEX TERMS Super-resolution reconstruction, pelvic, generative adversarial network, cyclic loss, attention.  ...  Therefore, we proposed a novel super-resolution method that uses a generative adversarial network with cyclic loss and attention mechanism to generate high-resolution MR images from low-resolution MR images  ...  The RED [11] used an encoder-decoder framework in the super-resolution method.  ... 
doi:10.1109/access.2021.3099695 fatcat:izcyarzbendp5ni64u2odr5tue

Joint Light Field Spatial and Angular Super-Resolution from a Single Image

Andre Ivan, Williem, In Kyu Park
2020 IEEE Access  
The conventional method relies on physical-based rendering and a secondary network to solve the angular super-resolution problem.  ...  In this paper, we show that both super-resolution problems can be solved jointly from a single image by proposing a single end-to-end deep neural network that does not require a physical-based approach  ...  The idea is to jointly synthesize light field through angular and spatial super-resolution (SR) from only a single image which is abundantly available in the real world.  ... 
doi:10.1109/access.2020.3002921 fatcat:xcxl6lp2hzfsljqeiufzmje3wi

Multi Scale Identity-Preserving Image-to-Image Translation Network for Low-Resolution Face Recognition [article]

Vahid Reza Khazaie and Nicky Bayat and Yalda Mohsenzadeh
2021 arXiv   pre-print
We achieved this by training a very deep convolutional encoder-decoder network with a symmetric contracting path between corresponding layers.  ...  Here, we propose an identity-preserving end-to-end image-to-image translation deep neural network which is capable of super-resolving very low-resolution faces to their high-resolution counterparts while  ...  [12] was to develop a residual neural network for superresolution that enhances gradient images.  ... 
arXiv:2010.12249v3 fatcat:dwjszfc6cvh5vmkuji35kerbxm

Light Field Image Compression via CNN-Based EPI Super-Resolution and Decoder-Side Quality Enhancement

Jinbo Zhao, Ping An, Xinpeng Huang, Chao Yang, Liquan Shen
2019 IEEE Access  
We propose a multi-scale dense residual network (MSDRN) to implement both EPI super-resolution and quality enhancement.  ...  The epipolar plane image (EPI) super-resolution is for compensating the information loss caused by sparse sampling and the decoder-side sub-aperture images (SAIs) quality enhancement is for compensating  ...  In order to implement both EPI super-resolution and decoder-side quality enhancement, we design a multi-scale dense residual network (MSDRN), which extracts features with multi-scale filters and passes  ... 
doi:10.1109/access.2019.2930644 fatcat:xjj63zrdrncltbhtmbvwp6hd2a

Video compression with low complexity CNN-based spatial resolution adaptation [article]

Di Ma, Fan Zhang, David R. Bull
2020 arXiv   pre-print
This approach employs a CNN model for video down-sampling at the encoder and uses a Lanczos3 filter to reconstruct full resolution at the decoder.  ...  However, this approach suffers from high complexity at the decoder due to the employment of CNN-based super-resolution.  ...  Since the proposed network is used for resolution downsampling before encoding, the CNN output (at low resolution) should preserve sufficient spatial information to enble high fidelity full resolution  ... 
arXiv:2007.14726v1 fatcat:vzjsjjcd6zhotgcqvxa4l7f74u

FSRNet: End-to-End Learning Face Super-Resolution with Facial Priors

Yu Chen, Ying Tai, Xiaoming Liu, Chunhua Shen, Jian Yang
2018 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition  
We present a novel deep end-to-end trainable Face Super-Resolution Network (FSRNet), which makes use of the geometry prior, i.e., facial landmark heatmaps and parsing maps, to superresolve very low-resolution  ...  To generate realistic faces, we also propose the Face Super-Resolution Generative Adversarial Network (FSRGAN) to incorporate the adversarial loss into FSRNet.  ...  Face Super-Resolution Network Overview of FSRNet Our basic FSRNet F consists of four parts: coarse SR network, fine SR encoder, prior estimation network and finally a fine SR decoder.  ... 
doi:10.1109/cvpr.2018.00264 dblp:conf/cvpr/ChenT0S018 fatcat:5dcj4k2bsbhnzoepefzgfqpnte

Perceptual Image Super-Resolution with Progressive Adversarial Network [article]

Lone Wong, Deli Zhao, Shaohua Wan, Bo Zhang
2020 arXiv   pre-print
To address this issue, we propose Progressive Adversarial Network (PAN) that is capable of coping with this difficulty for domain-specific image super-resolution.  ...  The low-level features in encoder can be transferred into decoder to enhance textural details with U-Net.  ...  These connections are useful to save insights from different abstraction levels and transfer them from the encoder to the decoder network.  ... 
arXiv:2003.03756v4 fatcat:dg32vyec5ndhrhmpin7kp4uhwi

RestoreDet: Degradation Equivariant Representation for Object Detection in Low Resolution Images [article]

Ziteng Cui, Yingying Zhu, Lin Gu, Guo-Jun Qi, Xiaoxiao Li, Peng Gao, Zenghui Zhang, Tatsuya Harada
2022 arXiv   pre-print
Image restoration algorithms such as super resolution (SR) are indispensable pre-processing modules for object detection in degraded images.  ...  Specifically, we learn this intrinsic visual structure by encoding and decoding the degradation transformation from a pair of original and randomly degraded images.  ...  By removing this layer, Lim et al. proposed enhanced deep super resolution network (EDSR) [38] that achieves the SOTA in 2017.  ... 
arXiv:2201.02314v1 fatcat:dhzdddkseffotjzwpehrhryocq

Lossless Image Compression through Super-Resolution [article]

Sheng Cao, Chao-Yuan Wu, Philipp Krähenbühl
2020 arXiv   pre-print
For lossless super-resolution, we predict the probability of a high-resolution image, conditioned on the low-resolution input, and use entropy coding to compress this super-resolution operator.  ...  We store a low resolution version of an image as raw pixels, followed by several iterations of lossless super-resolution.  ...  In the following section, we describe our network for one level of super-resolution and omit the superscript for simplicity.  ... 
arXiv:2004.02872v1 fatcat:ocwhfs5f55ht3o2dc3w3lfkcbe

Brain MRI Super-resolution Using 3D Dilated Convolutional Encoder-Decoder Network

Jinglong Du, Lulu Wang, Yulu Liu, Zexun Zhou, Zhongshi He, Yuanyuan Jia
2020 IEEE Access  
INDEX TERMS Magnetic resonance image, super-resolution reconstruction, dilated convolutional neural networks, encoder-decoder network.  ...  To address this issue, we propose a novel dilated convolutional encoder-decoder (DCED) network to improve the resolution of MRI.  ...  CONCLUSION In this paper, we propose to reconstruct high-resolution MRI through a dilated encoder-decoder network.  ... 
doi:10.1109/access.2020.2968395 fatcat:utlyrnabobazbnanxgqxutgtg4
« Previous Showing results 1 — 15 out of 3,791 results