A new image representation algorithm inspired by image submodality models, redundancy reduction, and learning in biological vision

Abstract

Abstract-We develop a new biologically motivated algorithm for representing natural images using successive projections into complementary subspaces. An image is first projected into an edge subspace spanned using an ICA basis adapted to natural images which captures the sharp features of an image like edges and curves. The residual image obtained after extraction of the sharp image features is approximated using a mixture of probabilistic principal component analyzers (MPPCA) model. The model is consistent with cellular, functional, information theoretic, and learning paradigms in visual pathway modeling. We demonstrate the efficiency of our model for representing different attributes of natural images like color and luminance. We compare the performance of our model in terms of quality of representation against commonly used basis, like the discrete cosine transform (DCT), independent component analysis (ICA), and principal components analysis (PCA), based on their entropies. Chrominance and luminance components of images are represented using codes having lower entropy than DCT, ICA, or PCA for similar visual quality. The model attains considerable simplification for learning from images by using a sparse independent code for representing edges and explicitly evaluating probabilities in the residual subspace.

Quantifying Nearness in Visual Spaces

“…An image feature model of perception was suggested by Mojsilovic et al (2002), where it was suggested that humans view or recall an image by its dominant colors only, and areas containing small, nondominant colors are averaged by the human visual system. Other examples of the term perception defined in the context of psychophysics have also been given (Wilson et al 1997;Dempere-Marco et al 2002;Kuo and Johnson 2002;Wandell et al 2002;Wang et al 2004;Balakrishnan et al 2005;Qamra et al 2005).…”

Biologically Inspired Model with Feature Selection for Target Recognition Using Biogeography-Based Optimization

“…Establishing an emulating cortex recognition system has always been a challenging issue. References [11][12][13] have investigated the biological feasibility of explaining aspects of higher-level visual processing. The aim of object recognition can be understood by representing objects as collections of view-specific features and taking the view that the basic recognition processes occur in a bottom-up way.…”

Book of abstracts

“…This last alternative can be implemented by approximation [10], image simplification at early stages [11], sparse representation of the scene [12] or hierarchical processing and resource optimization [13]. For systems running on a low power budget, energy efficiency is required, therefore power consumption needs to be minimized:…”