A Tour of Modern Image Filtering: New Insights and Methods, Both Practical and Theoretical

By Peyman Milanfar. IEEE Signal Processing Magazine, 2013

610 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Media Technology

View Paper: https://www.semanticscholar.org/paper/5620050230ba1b45689cfc5e3efa064219a66a2e

How researchers see this paper

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

In Computer Vision and Pattern Recognition

In Media Technology

Citations in the corpus

Fast Image Processing with Fully-Convolutional Networks (2017, background): “(Deep mathematical connections between these families of operators exist [57].”
Multi-Scale Boosted Dehazing Network With Dense Feature Fusion (2020, methodology): “The boosting strategy [44, 6, 51] is originally developed for image denoising by progressively refining the intermediate result from the previous iteration, and the error feedback mechanism, especially the back-projection technique [27, 12, 22], is designed for super-resolution to progressively reco”
Cascade Residual Learning: A Two-Stage Convolutional Neural Network for Stereo Matching (2017, background): “In general, low-level vision tasks, e.g. , denoising and de-blurring, can be improved with post-facto iterative reﬁne-ment [22], and disparity/ﬂow estimation is no exception [3].”
Trainable Nonlinear Reaction Diffusion: A Flexible Framework for Fast and Effective Image Restoration (2015, background): “There exists a huge amount of literature addressing the topic of image restoration problems, see for example [41] for a survey.”
MemNet: A Persistent Memory Network for Image Restoration (2017, background): “Image restoration [29] is a classical problem in low-level computer vision, which estimates an uncorrupted image from a noisy or blurry one.”
Self-Distillation Amplifies Regularization in Hilbert Space (2020, methodology): “Ideas similar to self-distillation have been used in areas besides modern machine learning but with different names such diffusion and boosting in both the statistics and image processing communities [22].”
Plug-and-Play Priors for Bright Field Electron Tomography and Sparse Interpolation (2015, methodology): “However, this approach is not immediately applicable to the incorporation of a wider variety of non-local denoising algorithms such as are described by Milanfar in [37]–[39].”
Image Restoration Using Very Deep Convolutional Encoder-Decoder Networks with Symmetric Skip Connections (2016, neutral): “See detail reviews in a survey [21].”
Graph Learning From Data Under Laplacian and Structural Constraints (2016, methodology): “video signals can be modeled using weighted graphs whose weights capture the correlation or similarity between neighboring pixel values (such as in nearest-neighbor models) [5]–[10].”
Graph Signal Processing: Overview, Challenges, and Applications (2017, methodology): “This type of image-dependent graph representation is strongly connected to popular image processing techniques, such as the bilateral filter and related methods [73], which also apply signal dependent filtering and are widerly used in applications such as image”
The Little Engine That Could: Regularization by Denoising (RED) (2016, neutral): “This definition by itself is not novel, as it is similar to ideas brought up in a series of recent contributions [47, 48, 49, 50, 51, 52, 53, 54, 55, 56].”