DOCSLIB.ORG

Generative Adversarial Networks for Single Image Super Resolution in Microscopy Images

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Generative Adversarial Networks for Single Image Super Resolution in Microscopy Images DEGREE PROJECT IN INFORMATION AND COMMUNICATION TECHNOLOGY, SECOND CYCLE, 30 CREDITS STOCKHOLM, SWEDEN 2018 Generative adversarial networks for single image super resolution in microscopy images SAURABH GAWANDE KTH ROYAL INSTITUTE OF TECHNOLOGY SCHOOL OF INFORMATION AND COMMUNICATION TECHNOLOGY Generative adversarial networks for single image super resolution in microscopy images SAURABH GAWANDE Master’s Thesis at KTH Information and Communication Technology Supervisor: Mihhail Matskin Examiner: Dr. Anne Hakånsson Industrial Supervisor: Dr. Kevin Smith TRITA-EECS-EX-2018:10 Abstract Image Super resolution is a widely-studied problem in computer vision, where the objective is to convert a low- resolution image to a high resolution image. Conventional methods for achieving super-resolution such as image pri- ors, interpolation, sparse coding require a lot of pre/post processing and optimization. Recently, deep learning meth- ods such as convolutional neural networks and generative adversarial networks are being used to perform super-resolution with results competitive to the state of the art but none of them have been used on microscopy images. In this thesis, a generative adversarial network, mSRGAN, is proposed for super resolution with a perceptual loss function consist- ing of a adversarial loss, mean squared error and content loss. The objective of our implementation is to learn an end to end mapping between the low / high resolution images and optimize the upscaled image for quantitative metrics as well as perceptual quality. We then compare our results with the current state of the art methods in super reso- lution, conduct a proof of concept segmentation study to show that super resolved images can be used as a effective pre processing step before segmentation and validate the findings statistically. Keywords: Deep Learning, Generative adversarial net- works, Super resolution, High content screening microscopy Abstract Image Super-resolution är ett allmänt studerad problem i datasyn, där målet är att konvertera en lågupplösnings- bild till en högupplöst bild. Konventionella metoder för att uppnå superupplösning som image priors, interpolation, sparse coding behöver mycket för- och efterbehandling och optimering.Nyligen djupa inlärningsmetoder som convolu- tional neurala nätverk och generativa adversariella nätverk är användas för att utföra superupplösning med resultat som är konkurrenskraftiga mot toppmoderna teknik, men ingen av dem har använts på mikroskopibilder. I denna avhandling, ett generativ kontradiktorisktsnätverk, mSR- GAN, är föreslås för superupplösning med en perceptuell förlustfunktion bestående av en motsatt förlust, medelk- vadratfel och innehållförlust.Mål med vår implementering är att lära oss ett slut på att slut kartläggning mellan bilder med låg / hög upplösning och optimera den uppskalade bilden för kvantitativa metriks såväl som perceptuell kvalitet. Vi jämför sedan våra resultat med de nuvarande toppmod- erna metoderna i superupplösning, och uppträdande ett be- vis på konceptsegmenteringsstudie för att visa att superlösa bilder kan användas som ett effektivt förbehandling steg före segmentering och validera fynden statistiskt. Keywords: Deep Learning, Generative adversarial net- works, Super resolution, High content screening microscopy Acknowledgements First of all, I would like to express my sincerest gratitude to Dr. Kevin Smith for giving me the opportunity to work on this exciting topic and without whom this work wouldn’t have materialized. Having a guide like Kevin was truly a blessing and I could not have wished for a better mentor. Thank you, Kevin for always being patient with me, pushing me to go the extra mile and always making the time for me despite your hectic schedule. I would like to thank Dr.Hossein Azizpour for his continuous feedback and ideas, always being available to clear my doubts no matter how naive and serving as a beacon of inspiration. I am also thankful to my examiner Dr. Anne Hakånsson for providing me the support I needed to stay on track and helping me maintaining scientific quality of this work . Last but not the least, no amount of thanks will ever be enough for my parents who have loved, supported and cared for me unconditionally throughout my tumul- tuous and protracted journey. May all your minima always be local! Tack! Contents Abbreviations 1 Introduction 1 1.1 Image Super-resolution . 1 1.2 Background . 2 1.3 Problem . 3 1.4 Purpose and Goal . 3 1.5 Ethics and Sustainability . 4 1.6 Methodology . 4 1.7 Delimitations . 5 1.8 Outline . 5 1.9 Contributions . 6 2 Relevant Theory 7 2.1 Background knowledge . 7 2.1.1 Definitions . 7 2.1.2 Strategies to increase image resolution . 9 2.1.3 Evaluation metric for Super-Resolution . 10 2.2 Neural Networks . 10 2.2.1 Convolutional Neural Networks . 13 2.2.2 Generative Adversarial Networks . 13 2.3 Literature Study . 17 2.3.1 Traditional Single Image super resolution . 18 2.3.2 Deep Learning based Single Image super resolution . 20 3 Motivation 24 3.1 HCS microscopy problems in image acquisition . 24 3.1.1 Photo bleaching . 25 3.1.2 Bleed through/ Crosstalk . 27 3.1.3 Phototoxicity . 28 3.1.4 Uneven illumination . 29 3.1.5 Color and contrast errors . 30 3.2 Inefficiency of pixel wise M.S.E . 31 3.3 Feature transferability issues in CNN’s for distant source and target domians . 33 4 Methods 35 4.1 Research Methods . 35 4.2 Mathematical formulation . 37 4.3 Generative Adversarial Network architecture . 38 4.4 Loss functions . 40 4.4.1 Perceptual Loss . 40 4.4.2 Pixel wise Mean squared error . 41 4.4.3 Content Loss . 42 4.4.4 Adversarial Loss . 42 4.4.5 Flowchart . 42 4.5 Data Acquisition . 44 4.5.1 Data processing . 46 5 Experiments and Results 47 5.1 mSRGAN . 49 5.2 mSRGAN - VGG2 . 52 5.3 mSRGAN - VGG5 . 54 5.4 SRRESNET . 57 5.5 mSRGAN - CL (Only content loss) . 60 5.6 SRGAN . 64 5.7 Nuclei Segmentation . 68 5.7.1 Statistical Validation . 70 6 Discussion 74 6.1 mSRGAN vs SRGAN (Evaluating Hypothesis 1) . 74 6.2 Content loss vs M.S.E (Evaluating hypothesis 2) . 76 6.2.1 Effect of Different VGG layers . 77 6.2.2 PSNR variation for mSRGAN variants . 78 6.3 Segmentation results . 79 6.4 GAN failures . 80 6.5 Checkerboard Artifacts . 91 6.6 Lack of training data . 92 7 Conclusion 93 7.1 Future Work . 94 Bibliography 95 Appendices 101 A More Results 101 Abbreviations α weight coefficient for M.S.E β weight coefficient for content loss CNN Convolutional neural networks CT Computed tomography DL Deep Learning GAN Generative adversarial networks GAN Generative adversarial networks HCS High content screening HR High resolution HVS Human visual system LR low resolution image MRI Magnetic resonance imaging MSE Mean Squared Error psnr Peak Signal to Noise ratio SC Sparse coding SGD Stochastic gradient descent SISR Single image super resolution SR Super Resolution Chapter 1 Introduction In this thesis project, we explore the use of Generative adversarial networks for per- forming single image super resolution on high content screening microscopy images. The project was carried out within the Bioimage Informatics Facility at the Science for Life Laboratory, Sweden. 1.1 Image Super-resolution In most digital imaging applications, high-resolution images are preferred and of- ten required to accomplish tasks. Image super-resolution (SR) is a widely-studied problem in computer vision, where the objective is to generate one or more high- resolution images from one or more low-resolution images. SR algorithm aims to produce details finer than the sampling grid of a given imaging device by increas- ing the number of pixels per unit area in an image. SR is a well known ill-posed inverse problem, where from a low-resolution image (usually corrupted by noise, mo- tion blur, aliasing, optical distortion, etc.) a high-resolution image is restored [1] [2]. SR techniques can be applied in many scenarios where multiple frames of a sin- gle scene can be obtained (e.g., multiple images of the same object by a single camera), various images of a scene are available from numerous sources (numerous cameras capturing a single scene from various locations). SR has its applications in varied fields such as Satellite imaging (eg. remote sens- ing) where several images of a single area are available, in security and surveillance where it may be required to enlarge a particular point of interest in a scene (such as zooming on the face of a criminal or the numbers of a license plate), in computer vision where it can improve the performance of pattern recognition and other areas such as facial image analysis, text image analysis, biometric identification, finger- print image enhancement, etc. [1]. SR is particularly of great importance in medical imaging where more detailed 1 CHAPTER 1. INTRODUCTION image details are required on demand, and high-resolution medical images can aid the doctors to make a correct diagnosis, e.g., in Computed tomography (CT) and Magnetic resonance imaging (MRI) for diagnosis, where the acquisition of multiple images is possible albeit with limited resolution. 1.2 Background Convolutional neural networks (CNN) have been in existence for a long time [2] and recently deep CNN’s have shown an upsurge in popularity due to its various successes in image classification tasks, one of them being the ImageNet Large Scale Visual Recognition Challenge which is a benchmark in object classification and de- tection tasks consisting of millions of images and thousands of classes [3]. CNN’s have also been applied to other sub-problems of computer vision such as object detection [4], face recognition [5] and pedestrian detection [6].Various factors are instrumental in the progress and effectiveness of CNN’s such as A) The advent of more powerful Graphics Processing Units [3], which makes it easier to train more complex models on large datasets B) The exponential increase in the amount of big data which helps in training large models and getting more accurate results.
Load more

Recommended publications
  • Multiframe Motion Coupling for Video Super Resolution
    Multiframe Motion Coupling for Video Super Resolution Jonas Geiping,* Hendrik Dirks,† Daniel Cremers,‡ Michael Moeller§ December 5, 2017 Abstract The idea of video super resolution is to use different view points of a single scene to enhance the overall resolution and quality. Classical en- ergy minimization approaches first establish a correspondence of the current frame to all its neighbors in some radius and then use this temporal infor- mation for enhancement. In this paper, we propose the first variational su- per resolution approach that computes several super resolved frames in one batch optimization procedure by incorporating motion information between the high-resolution image frames themselves. As a consequence, the number of motion estimation problems grows linearly in the number of frames, op- posed to a quadratic growth of classical methods and temporal consistency is enforced naturally. We use infimal convolution regularization as well as an automatic pa- rameter balancing scheme to automatically determine the reliability of the motion information and reweight the regularization locally. We demonstrate that our approach yields state-of-the-art results and even is competitive with machine learning approaches. arXiv:1611.07767v2 [cs.CV] 4 Dec 2017 1 Introduction The technique of video super resolution combines the spatial information from several low resolution frames of the same scene to produce a high resolution video. *University of Siegen, [email protected] †University of Munster,¨ [email protected] ‡Technical University of Munich, [email protected] §University of Siegen, [email protected] 1 A classical way of solving the super resolution problem is to estimate the motion from the current frame to its neighboring frames, model the data formation process via warping, blur, and downsampling, and use a suitable regularization to suppress possible artifacts arising from the ill-posedness of the underlying problem.
    [Show full text]
  • Video Super-Resolution Based on Spatial-Temporal Recurrent Residual Networks
    1 Computer Vision and Image Understanding journal homepage: www.elsevier.com Video Super-Resolution Based on Spatial-Temporal Recurrent Residual Networks Wenhan Yanga, Jiashi Fengb, Guosen Xiec, Jiaying Liua,∗∗, Zongming Guoa, Shuicheng Yand aInstitute of Computer Science and Technology, Peking University, Beijing, P.R.China, 100871 bDepartment of Electrical and Computer Engineering, National University of Singapore, Singapore, 117583 cNLPR, Institute of Automation, Chinese Academy of Sciences, Beijing, P.R.China, 100190 dArtificial Intelligence Institute, Qihoo 360 Technology Company, Ltd., Beijing, P.R.China, 100015 ABSTRACT In this paper, we propose a new video Super-Resolution (SR) method by jointly modeling intra-frame redundancy and inter-frame motion context in a unified deep network. Different from conventional methods, the proposed Spatial-Temporal Recurrent Residual Network (STR-ResNet) investigates both spatial and temporal residues, which are represented by the difference between a high resolution (HR) frame and its corresponding low resolution (LR) frame and the difference between adjacent HR frames, respectively. This spatial-temporal residual learning model is then utilized to connect the intra-frame and inter-frame redundancies within video sequences in a recurrent convolutional network and to predict HR temporal residues in the penultimate layer as guidance to benefit esti- mating the spatial residue for video SR. Extensive experiments have demonstrated that the proposed STR-ResNet is able to efficiently reconstruct videos with diversified contents and complex motions, which outperforms the existing video SR approaches and offers new state-of-the-art performances on benchmark datasets. Keywords: Spatial residue, temporal residue, video super-resolution, inter-frame motion con- text, intra-frame redundancy.
    [Show full text]
  • Online Multi-Frame Super-Resolution of Image Sequences Jieping Xu1* , Yonghui Liang2,Jinliu2, Zongfu Huang2 and Xuewen Liu2
    Xu et al. EURASIP Journal on Image and Video EURASIP Journal on Image Processing (2018) 2018:136 https://doi.org/10.1186/s13640-018-0376-5 and Video Processing RESEARCH Open Access Online multi-frame super-resolution of image sequences Jieping Xu1* , Yonghui Liang2,JinLiu2, Zongfu Huang2 and Xuewen Liu2 Abstract Multi-frame super-resolution recovers a high-resolution (HR) image from a sequence of low-resolution (LR) images. In this paper, we propose an algorithm that performs multi-frame super-resolution in an online fashion. This algorithm processes only one low-resolution image at a time instead of co-processing all LR images which is adopted by state- of-the-art super-resolution techniques. Our algorithm is very fast and memory efficient, and simple to implement. In addition, we employ a noise-adaptive parameter in the classical steepest gradient optimization method to avoid noise amplification and overfitting LR images. Experiments with simulated and real-image sequences yield promising results. Keywords: Super-resolution, Image processing, Image sequences, Online image processing 1 Introduction There are a variety of SR techniques, including multi- Images with higher resolution are required in most elec- frame SR and single-frame SR. Readers can refer to Refs. tronic imaging applications such as remote sensing, med- [1, 5, 6] for an overview of this issue. Our discussion below ical diagnostics, and video surveillance. For the past is limited to work related to fast multi-frame SR method, decades, considerable advancement has been realized in as it is the focus of our paper. imaging system. However, the quality of images is still lim- One SR category close to our alogorithm is dynamic SR ited by the cost and manufacturing technology [1].
    [Show full text]
  • Versatile Video Coding and Super-Resolution For
    VERSATILE VIDEO CODING AND SUPER-RESOLUTION FOR EFFICIENT DELIVERY OF 8K VIDEO WITH 4K BACKWARD-COMPATIBILITY Charles Bonnineau, Wassim Hamidouche, Jean-Francois Travers, Olivier Deforges To cite this version: Charles Bonnineau, Wassim Hamidouche, Jean-Francois Travers, Olivier Deforges. VERSATILE VIDEO CODING AND SUPER-RESOLUTION FOR EFFICIENT DELIVERY OF 8K VIDEO WITH 4K BACKWARD-COMPATIBILITY. International Conference on Acoustics, Speech, and Sig- nal Processing (ICASSP), May 2020, Barcelone, Spain. hal-02480680 HAL Id: hal-02480680 https://hal.archives-ouvertes.fr/hal-02480680 Submitted on 16 Feb 2020 HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. VERSATILE VIDEO CODING AND SUPER-RESOLUTION FOR EFFICIENT DELIVERY OF 8K VIDEO WITH 4K BACKWARD-COMPATIBILITY Charles Bonnineau?yz, Wassim Hamidouche?z, Jean-Franc¸ois Traversy and Olivier Deforges´ z ? IRT b<>com, Cesson-Sevigne, France, yTDF, Cesson-Sevigne, France, zUniv Rennes, INSA Rennes, CNRS, IETR - UMR 6164, Rennes, France ABSTRACT vary depending on the exploited transmission network, reducing the In this paper, we propose, through an objective study, to com- number of use-cases that can be covered using this approach. For pare and evaluate the performance of different coding approaches instance, in the context of terrestrial transmission, the bandwidth is allowing the delivery of an 8K video signal with 4K backward- limited in the range 30-40Mb/s using practical DVB-T2 [7] channels compatibility on broadcast networks.
    [Show full text]
  • On Bayesian Adaptive Video Super Resolution
    IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE 1 On Bayesian Adaptive Video Super Resolution Ce Liu Member, IEEE, Deqing Sun Member, IEEE Abstract—Although multi-frame super resolution has been extensively studied in past decades, super resolving real-world video sequences still remains challenging. In existing systems, either the motion models are oversimplified, or important factors such as blur kernel and noise level are assumed to be known. Such models cannot capture the intrinsic characteristics that may differ from one sequence to another. In this paper, we propose a Bayesian approach to adaptive video super resolution via simultaneously estimating underlying motion, blur kernel and noise level while reconstructing the original high-res frames. As a result, our system not only produces very promising super resolution results outperforming the state of the art, but also adapts to a variety of noise levels and blur kernels. To further analyze the effect of noise and blur kernel, we perform a two-step analysis using the Cramer-Rao bounds. We study how blur kernel and noise influence motion estimation with aliasing signals, how noise affects super resolution with perfect motion, and finally how blur kernel and noise influence super resolution with unknown motion. Our analysis results confirm empirical observations, in particular that an intermediate size blur kernel achieves the optimal image reconstruction results. Index Terms—Super resolution, optical flow, blur kernel, noise level, aliasing F 1 INTRODUCTION image reconstruction, optical flow, noise level and blur ker- nel estimation. Using a sparsity prior for the high-res image, Multi-frame super resolution, namely estimating the high- flow fields and blur kernel, we show that super resolution res frames from a low-res sequence, is one of the fundamen- computation is reduced to each component problem when tal problems in computer vision and has been extensively other factors are known, and the MAP inference iterates studied for decades.
    [Show full text]
  • Video Super-Resolution Via Deep Draft-Ensemble Learning
    Video Super-Resolution via Deep Draft-Ensemble Learning Renjie Liao† Xin Tao† Ruiyu Li† Ziyang Ma§ Jiaya Jia† † The Chinese University of Hong Kong § University of Chinese Academy of Sciences http://www.cse.cuhk.edu.hk/leojia/projects/DeepSR/ Abstract Difficulties It has been observed for long time that mo- tion estimation critically affects MFSR. Erroneous motion We propose a new direction for fast video super- inevitably distorts local structure and misleads final recon- resolution (VideoSR) via a SR draft ensemble, which is de- struction. Albeit essential, in natural video sequences, suf- fined as the set of high-resolution patch candidates before ficiently high quality motion estimation is not easy to ob- final image deconvolution. Our method contains two main tain even with state-of-the-art optical flow methods. On components – i.e., SR draft ensemble generation and its op- the other hand, the reconstruction and point spread func- timal reconstruction. The first component is to renovate tion (PSF) kernel estimation steps could introduce visual traditional feedforward reconstruction pipeline and greatly artifacts given any errors produced in this process or the enhance its ability to compute different super-resolution re- low-resolution information is not enough locally. sults considering large motion variation and possible er- These difficulties make methods finely modeling all con- rors arising in this process. Then we combine SR drafts straints in e.g., [12], generatively involve the process with through the nonlinear process in a deep convolutional neu- sparse regularization for iteratively updating variables – ral network (CNN). We analyze why this framework is pro- each step needs to solve a set of nonlinear functions.
    [Show full text]
  • A Bayesian Approach to Adaptive Video Super Resolution
    A Bayesian Approach to Adaptive Video Super Resolution Ce Liu Deqing Sun Microsoft Research New England Brown University (a) Input low-res (b) Bicubic up-sampling ×4 (c) Output from our system (d) Original frame Figure 1. Our video super resolution system is able to recover image details after×4 up-sampling. trary, the video may be contaminated with noise of unknown Abstract level, and motion blur and point spread functions can lead Although multi-frame super resolution has been exten- to an unknown blur kernel. sively studied in past decades, super resolving real-world Therefore, a practical super resolution system should si- video sequences still remains challenging. In existing sys- multaneously estimate optical flow [9], noise level [18] and tems, either the motion models are oversimplified, or impor- blur kernel [12] in addition to reconstructing the high-res tant factors such as blur kernel and noise level are assumed frames. As each of these problems has been well studied in to be known. Such models cannot deal with the scene and computer vision, it is natural to combine all these compo- imaging conditions that vary from one sequence to another. nents in a single framework without making oversimplified In this paper, we propose a Bayesian approach to adaptive assumptions. video super resolution via simultaneously estimating under- In this paper, we propose a Bayesian framework for lying motion, blur kernel and noise level while reconstruct- adaptive video super resolution that incorporates high-res ing the original high-res frames. As a result, our system not image reconstruction, optical flow, noise level and blur ker- only produces very promising super resolution results that nel estimation.
    [Show full text]
  • Video Super-Resolution Based on Generative Adversarial Network and Edge Enhancement
    electronics Article Video Super-Resolution Based on Generative Adversarial Network and Edge Enhancement Jialu Wang , Guowei Teng * and Ping An Department of Signal and Information Processing, School of Communication and Information Engineering, Shanghai University, Shanghai 200444, China; [email protected] (J.W.); [email protected] (P.A.) * Correspondence: [email protected]; Tel.: +86-021-6613-5051 Abstract: With the help of deep neural networks, video super-resolution (VSR) has made a huge breakthrough. However, these deep learning-based methods are rarely used in specific situations. In addition, training sets may not be suitable because many methods only assume that under ideal circumstances, low-resolution (LR) datasets are downgraded from high-resolution (HR) datasets in a fixed manner. In this paper, we proposed a model based on Generative Adversarial Network (GAN) and edge enhancement to perform super-resolution (SR) reconstruction for LR and blur videos, such as closed-circuit television (CCTV). The adversarial loss allows discriminators to be trained to distinguish between SR frames and ground truth (GT) frames, which is helpful to produce realistic and highly detailed results. The edge enhancement function uses the Laplacian edge module to perform edge enhancement on the intermediate result, which helps further improve the final results. In addition, we add the perceptual loss to the loss function to obtain a higher visual experience. At the same time, we also tried training network on different datasets. A large number of experiments show that our method has advantages in the Vid4 dataset and other LR videos. Citation: Wang, J.; Teng, G.; An, P. Keywords: video super-resolution; generative adversarial networks; edge enhancement Video Super-Resolution Based on Generative Adversarial Network and Edge Enhancement.
    [Show full text]
  • Learning to Have an Ear for Face Super-Resolution
    Learning to Have an Ear for Face Super-Resolution Givi Meishvili Simon Jenni Paolo Favaro University of Bern, Switzerland {givi.meishvili, simon.jenni, paolo.favaro}@inf.unibe.ch Abstract We propose a novel method to use both audio and a low- resolution image to perform extreme face super-resolution (a 16× increase of the input size). When the resolution of the input image is very low (e.g., 8 × 8 pixels), the loss of information is so dire that important details of the origi- nal identity have been lost and audio can aid the recovery (a) (b) (c) (d) (e) (f) of a plausible high-resolution image. In fact, audio car- Figure 1: Audio helps image super-resolution. (a) and ries information about facial attributes, such as gender and (b) are the ground-truth and 16× downsampled images re- age. To combine the aural and visual modalities, we pro- spectively; (c) results of the SotA super-resolution method pose a method to first build the latent representations of a of Huang et al. [22]; (d) our super-resolution from only the face from the lone audio track and then from the lone low- low-res image; (e) audio only super-resolution; (f) fusion of resolution image. We then train a network to fuse these two both the low-res image and audio. In these cases all meth- representations. We show experimentally that audio can ods fail to restore the correct gender without audio. assist in recovering attributes such as the gender, the age and the identity, and thus improve the correctness of the age might be incorrect (see Fig.
    [Show full text]
  • Video Super Resolution Based on Deep Learning: a Comprehensive Survey
    1 Video Super Resolution Based on Deep Learning: A Comprehensive Survey Hongying Liu, Member, IEEE, Zhubo Ruan, Peng Zhao, Chao Dong, Member, IEEE, Fanhua Shang, Senior Member, IEEE, Yuanyuan Liu, Member, IEEE, Linlin Yang Abstract—In recent years, deep learning has made great multiple successive images/frames at a time so as to uti- progress in many fields such as image recognition, natural lize relationship within frames to super-resolve the target language processing, speech recognition and video super- frame. In a broad sense, video super-resolution (VSR) resolution. In this survey, we comprehensively investigate 33 state-of-the-art video super-resolution (VSR) methods based can be regarded as a class of image super-resolution on deep learning. It is well known that the leverage of and is able to be processed by image super-resolution information within video frames is important for video algorithms frame by frame. However, the SR perfor- super-resolution. Thus we propose a taxonomy and classify mance is always not satisfactory as artifacts and jams the methods into six sub-categories according to the ways of may be brought in, which causes unguaranteed temporal utilizing inter-frame information. Moreover, the architectures and implementation details of all the methods are depicted in coherence within frames. detail. Finally, we summarize and compare the performance In recent years, many video super-resolution algo- of the representative VSR method on some benchmark rithms have been proposed. They mainly fall into two datasets. We also discuss some challenges, which need to categories: the traditional methods and deep learning be further addressed by researchers in the community of based methods.
    [Show full text]
  • Image Super-Resolution: the Techniques, Applications, and Future
    Signal Processing 128 (2016) 389–408 Contents lists available at ScienceDirect Signal Processing journal homepage: www.elsevier.com/locate/sigpro Review Image super-resolution: The techniques, applications, and future Linwei Yue a, Huanfeng Shen b,c,n, Jie Li a, Qiangqiang Yuan c,d, Hongyan Zhang a,c, Liangpei Zhang a,c,n a The State Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan University, Wuhan, PR China b School of Resource and Environmental Science, Wuhan University, Wuhan, PR China c Collaborative Innovation Center of Geospatial Technology, Wuhan University, Wuhan, PR China d School of Geodesy and Geomatics, Wuhan University, Wuhan, PR China article info abstract Article history: Super-resolution (SR) technique reconstructs a higher-resolution image or sequence from the observed Received 8 October 2015 LR images. As SR has been developed for more than three decades, both multi-frame and single-frame SR Received in revised form have significant applications in our daily life. This paper aims to provide a review of SR from the per- 1 March 2016 spective of techniques and applications, and especially the main contributions in recent years. Reg- Accepted 1 May 2016 ularized SR methods are most commonly employed in the last decade. Technical details are discussed in Available online 14 May 2016 this article, including reconstruction models, parameter selection methods, optimization algorithms and Keywords: acceleration strategies. Moreover, an exhaustive summary of the current applications using SR techni- Super resolution ques has been presented. Lastly, the article discusses the current obstacles for future research. Resolution enhancement & 2016 Elsevier B.V.
    [Show full text]
  • Arxiv:2103.10081V1 [Cs.CV] 18 Mar 2021 Trained VSR Networks to Generate Training Targets for the Imaging, and Electronics (E.G., Smartphones, TV)
    Self-Supervised Adaptation for Video Super-Resolution Jinsu Yoo and Tae Hyun Kim Hanyang University, Seoul, Korea fjinsuyoo,[email protected] Abstract have studied “zero-shot” SR approaches, which exploit sim- ilar patches across different scales within the input image Recent single-image super-resolution (SISR) networks, and video [9, 13, 33, 32]. However, searching for LR-HR which can adapt their network parameters to specific input pairs of patches within LR images is also difficult. More- images, have shown promising results by exploiting the in- over, the number of self-similar patches critically decreases formation available within the input data as well as large as the scale factor increases [41], and the searching problem external datasets. However, the extension of these self- becomes more challenging. To improve the performance supervised SISR approaches to video handling has yet to by easing the searching phase, a coarse-to-fine approach is be studied. Thus, we present a new learning algorithm widely used [13, 33]. Recently, several neural approaches that allows conventional video super-resolution (VSR) net- that utilize external and internal datasets have been intro- works to adapt their parameters to test video frames with- duced and have produced satisfactory results [28, 34, 21]. out using the ground-truth datasets. By utilizing many self- However, these methods remain bounded to the smaller similar patches across space and time, we improve the per- scaling factor possibly because of a conventional approach formance of fully pre-trained VSR networks and produce to self-supervised data pair generation during the test phase. temporally consistent video frames.
    [Show full text]