The Design and Use of Steerable Filters

By William T. Freeman, Edward H. Adelson. 1991

Subfields: Computer Vision and Pattern Recognition, Signal Processing

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 Signal Processing

Citations in the corpus

Fast Edge Detection Using Structured Forests (2014, background): “The detection of edges is a critical preprocessing step for a variety of tasks, including object recognition [47], [17], segmentation [33], [1], and active contours [26].”
Deformable Convolutional Networks (2017, background): “Under certain conditions, such filters form a set of basis and their linear combination forms new filters within the same group of geometric transformations, such as multiple orientations in Steerable Filters [12] and multiple scales in [43].”
Invariant Representations without Adversarial Training (2018, background): “Early examples include Steerable Filters [8, 11], and later SIFT [16].”
Learning-based Video Motion Magnification (2018, methodology): “[1,6] use the complex spatial pyramid [7] to extract a phase-based representation.”
Covariant Compositional Networks For Learning Graphs (2018, background): “This phenomenon is called steerability , and has a signiﬁcant literature in both classical signal processing (Freeman & Adelson, 1991; Simoncelli et al., 1992; Perona, 1995; Teo & Hel-Or, 1998; Manduchi et al., 1998) and the neural networks ﬁeld (Cohen & Welling, 2016b).”
RotDCF: Decomposition of Convolutional Filters for Rotation-Equivariant Deep Networks (2018, background): “This typically incurs a significant increase in the number of parameters and computational load, even with the adoption of steerable filters [7, 36, 37].”
Harmonic Networks: Deep Translation and Rotation Equivariance (2016, background): “Steerable filters At the root of H-Nets lies the property of filter steerability [7].”
Understanding Image Representations by Measuring Their Equivariance and Equivalence (2014, background): “The equivariance maps and steerable filters (Freeman and Adelson 1991) share some of the underlying theory.”
Unsupervised Learning of Edges (2015, methodology): “Edge Detection: Early edge detectors were manually designed to use image gradients [14, 8, 15] and later texture gradients [2].”
Learning Steerable Filters for Rotation Equivariant CNNs (2017, methodology): “This definition includes the classical formulation of steerability by Freeman and Adelson [7] as a specific choice of basis.”
Polar Transformer Networks (2017, methodology): “An exact steerability framework was started in [6] where rotational steerability for Gaussian derivatives was explicitly computed, and was extended to the shiftable pyramid [24] handling rotations and scale.”
CubeNet: Equivariance to 3D Rotation and Translation (2018, background): “In the presence of continuous transformations, Freeman and Adelson famously [17] (and less famously Lenz [32]), shored up the theory of steerable filters, which are a set of bandlimited linear filters wθ∈R H×W , which can be synthesized exactly at any rotation θ as a finite linear combination of bas”
On the "steerability" of generative adversarial networks (2019, methodology): “We further show the 1We use the term “steerable” in analogy to the classic steerable filters of Freeman & Adelson [7].”
Oriented edge forests for boundary detection (2014, background): “Early work such as the Canny edge detector [6] focused on detecting brightness edges, estimating their orientation [11] and analyzing the theoretical limits of detection in the presence of image noise.”