Discriminative Sparse Image Models for Class-Specific Edge Detection and Image Interpretation

By Julien Mairal, Marius Leordeanu, Francis R. Bach, Martial Hebert, Jean Ponce. 2008

239 citations recorded by the source corpus.

Subfields: 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 Computer Vision and Pattern Recognition

Citations in the corpus

Fast Edge Detection Using Structured Forests (2014, background): “Unfortunately, many visually salient edges do not correspond to color gradients, such as texture edges [34] and illusory contours [39].”
Visual Boundary Prediction: A Deep Neural Prediction Network and Quality Dissection (2014, methodology): “(5) Finally, as in several previous works [Doll´ar and Zitnick, 2013, Mairal et al., 2008], we apply non-maximum suppression by the Canny method to the network predictions.”
Image Inpainting by Patch Propagation Using Patch Sparsity (2010, methodology): “This framework has been widely applied to image denoising [29], [30], edge detection [31], recognition [32], [33], super-resolution [34], texture synthesis [35], etc.”
Heterogeneous feature machines for visual recognition (2009, background): “In addition, the work in [33] [23] considers only a single type of features, while our model aims to efficiently fuse heterogeneous features.”
Discriminatively Trained Sparse Code Gradients for Contour Detection (2012, methodology): “…the availability of datasets with human-marked groundtruth contours, a variety of approaches have been proposed and evaluated (see a summary in [2]), such as learning to classify [17] [20] [16], contour grouping [23] [31] [12], multi-scale features [21] [2], and hierarchical region analysis [2].”
Sketch Tokens: A Learned Mid-level Representation for Contour and Object Detection (2013, background): “While numerous detection approaches [5, 11, 7] utilize such features directly, other papers learn edge-based features [27, 38, 21] or class-speciﬁc edges [23, 18] using object level supervision.”
Sparse Modeling for Image and Vision Processing (2014, methodology): “To deal with unobserved information, the dictionary learning formulation can be modified by introducing a binary mask Mi for every patch indexed by i [Mairal et al., 2008c,e].”
Sparse Representation Based Fisher Discrimination Dictionary Learning for Image Classification (2014, methodology): “(16) can be efficiently solved by updating each dictionary atom one by one via the algorithm like (Yang et al. 2010c) or (Mairal et al. 2008c).”
Semantic Image Segmentation with Contextual Hierarchical Models (2015, methodology): “[56] proposed to learn discriminative sparse dictionaries to distinguish between “patches centered on an edge pixel” and “patches centered on a non-edge pixel”.”
Boundary Detection Using F-Measure-, Filter- and Feature- (F3) Boost (2010, background): “This approach has led to improved detectors based on Boosting [6], topological properties of the image [7], spectral gradients [8], multiscale processing [9] and sparse dictionaries [10], among others.”
Shape Sharing for Object Segmentation (2012, background): “In contrast, our approach leverages shape in a category-independent manner, and thus does not require prior knowledge about the object(s) present.”
Toward Holistic Scene Understanding: Feedback Enabled Cascaded Classification Models (2012, methodology): “We do this by introducing the l1 sparsity in the parameters [56].”
Sparse Representation for Computer Vision and Pattern Recognition (2010, methodology): “This idea was then applied in [50] for learning to detect edges as part of an image classification system.”
Energy Disaggregation via Discriminative Sparse Coding (2010, methodology): “The algorithm is similar in spirit to a number of recent approaches to discrim inative training of sparse representations [12, 17, 18].”
New Algorithms for Learning Incoherent and Overcomplete Dictionaries (2013, methodology): “In image processing such discovered dictionaries are used to perform denoising [21], edge-detection [37], super-resolution [49] and compression.”