Recursive ICA

By Honghao Shan, Lingyun Zhang, Garrison W. Cottrell. 2006

25 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Cognitive Neuroscience, 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 Artificial Intelligence

1. Karl Ridgeway
2. Barbara Olshausen
3. Siwei Lyu
4. Charles F. Cadieu
5. Eero P. Simoncelli
6. Garrison W. Cottrell
7. Honghao Shan
8. Tim K. Marks

In Cognitive Neuroscience

1. Siwei Lyu
2. Barbara Olshausen
3. Charles F. Cadieu
4. Karl Ridgeway
5. Eero P. Simoncelli
6. Garrison W. Cottrell
7. Honghao Shan
8. Lingyun Zhang

In Signal Processing

1. Karl Ridgeway
2. Barbara Olshausen
3. Charles F. Cadieu
4. Siwei Lyu
5. Eero P. Simoncelli
6. Garrison W. Cottrell
7. Honghao Shan
8. Lingyun Zhang

Citations in the corpus

• A Survey of Inductive Biases for Factorial Representation-Learning (2016, methodology): “An R-ICA representation is less restrictive than PCA and ICA; it enables the discovery of more abstract factors.”
• Learning Intermediate-Level Representations of Form and Motion from Natural Movies (2012, methodology): “Note that transforming the marginal distributions of coefficients has been utilized by a number of image modeling researchers (Chen and Gopinath, 2000; Shan et al., 2007; Lyu and Simoncelli, 2009) and has an interesting relation to divisive normalization (Simoncelli, 1997; Schwartz and Simoncelli, 2”
• Efficient Visual Coding: From Retina To V2 (2013, methodology): “ceptive fields of V1 simple cells. Here we propose a hierarchical model following the efficient coding principle, as illustrated in Figure 1. This model is a variant on the Recursive ICA model (RICA) [1]. The model has two characteristics. First, this model adopts a layered framework: we apply the e”
• Nonlinear Extraction of Independent Components of Natural Images Using Radial Gaussianization (2009, background): “…subspace analysis (Hyv¨arinen & Hoyer, 2000), topographical ICA (Hyv¨arinen et al., 2001), and hierarchical ICA models (Karklin & Lewicki, 2005; Shan, Zhang, & Cottrell, 2007) all assume that image data are generated from linearly transformed densities that are formed by combining clusters of…”
• SUN: A Bayesian framework for saliency using natural statistics. (2008, methodology): “Whereas linear ICA features learned from natural images respond to discontinuities in illumination or color, higher-order non-linear ICA features are found to respond to discontinuity in textures (Karklin & Lewicki, 2003; Osindero, Welling, & Hinton, 2006; Shan, Zhang, & Cottrell, 2007).”