Image Super-Resolution Via Sparse Representation

By Jianchao Yang, John Wright, Thomas S. Huang, Yi Ma. 2010

5,261 citations recorded by the source corpus.

Subfields: Computational Mechanics, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computational Mechanics

In Computer Vision and Pattern Recognition

Citations in the corpus

Coupled Generative Adversarial Networks (2016, methodology): “The same limitation is also applied to dictionary learning-based approaches [2, 4].”
Perceptual Losses for Real-Time Style Transfer and Super-Resolution (2016, methodology): “They group super-resolution techniques into prediction-based methods (bilinear, bicubic, Lanczos, [29]), edgebased methods [30,31], statistical methods [32–34], patch-based methods [30,35– 41], and sparse dictionary methods [42,43].”
Amortised MAP Inference for Image Super-resolution (2016, methodology): “The most direct approach to MAP inference uses a model of the natural image prior and finds the MAP solution via gradient descent for each LR image (Datsenko & Elad, 2007; Yang et al., 2010).”
Multi-View Silhouette and Depth Decomposition for High Resolution 3D Object Representation (2018, methodology): “Traditionally, image super-resolution has used dictionary-style methods [7, 48], matching patches of images to higher-resolution counterparts.”
Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network (2016, methodology): “This approach is proposed in the methods of [47, 8].”
Image Super-Resolution Using Deep Convolutional Networks (2014, methodology): “These methods either exploit internal similarities of the same image [5], [13], [16], [19], [47], or learn mapping functions from external low- and high-resolution exemplar pairs [2], [4], [6], [15], [23], [25], [37], [41], [42], [47], [48], [50], [51].”
HOGgles: Visualizing Object Detection Features (2013, methodology): “To do this, we solve a paired dictionary learning problem, inspired by recent super resolution sparse coding work [23, 21]:”
EnhanceNet: Single Image Super-Resolution Through Automated Texture Synthesis (2016, neutral): “They further include dictionary-based methods [33, 40, 52, 59, 61, 64] that learn a sparse representation of image patches as a combination of dictionary atoms, as well as neural network-based approaches [4, 8, 9, 24, 25, 26, 47, 48, 62] which apply convolutional neural networks (CNNs) to the task…”
Accurate Image Super-Resolution Using Very Deep Convolutional Networks (2015, methodology): “For example, RFL [18] has two methods, where the ﬁrst one uses 91 images from Yang et al. [25] and the second one uses 291 images with the addition of 200 images from Berkeley Segmentation Dataset [16].”
Deeply-Recursive Convolutional Network for Image Super-Resolution (2015, methodology): “For training, we use 91 images proposed in Yang et al. [31] for all experiments.”
YOLACT: Real-Time Instance Segmentation (2019, background): “Classical representations include textons [23] and visual words [40], with advances made via sparsity and locality priors [44, 43, 46].”
Enhanced Deep Residual Networks for Single Image Super-Resolution (2017, methodology): “Those learning meth-ods rely on techniques ranging from neighbor embedding [3, 2, 7, 21] to sparse coding [31, 32, 27, 33].”
Deep Laplacian Pyramid Networks for Fast and Accurate Super-Resolution (2017, methodology): “Numerous SR meth-ods learn the LR-HR mapping with image pairs collected from external databases using supervised learning algorithms, such as nearest neighbor [10], manifold embedding [2, 5], kernel ridge regression [19], and sparse representation [37, 38, 39].”
Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising (2016, methodology): “To learn a single model for the three general image de-noising tasks, as in [35], we use a dataset which consists of 91 images from [36] and 200 training images from the Berkeley segmentation dataset.”
Learning a No-Reference Quality Metric for Single-Image Super-Resolution (2016, methodology): “(d) Glasner09 [7] (e) Yang10 [8] (f) Dong11 [9]”
Training-Free, Single-Image Super-Resolution Using a Dynamic Convolutional Network (2018, background): “[6], [7], who trained dictionaries for LR and HR image patches.”
Neural Nearest Neighbors Networks (2018, methodology): “n a 80 80 region. Training details for super-resolution. We follow the training protocol of [20]. Our training set consists of 291 images: The 200 images of the BSD500 training set and 91 images from [48]. In each of the 80 training epochs, we randomly crop 3833 patches of size 80 80 from each image”