Biological Vision Systems Research Articles

A Chaos Theoretic Analysis of Motion and Illumination in Video Sequences

Accurate and robust image motion detection has been of substantial interest in the image processing and computer vision communities. Unfortunately, no single motion detection algorithm has been universally superior, yet biological vision systems are adept at motion detection. Recent research in neural signals have shown biological neural systems are highly responsive to signals which appear to be chaotic in nature. In this paper, we exploit these biological results and hypothesize that motion in images may produce changes in pixel amplitudes that are reminiscent of chaotic dynamical systems. In particular, we demonstrate that the trajectories of pixel amplitudes in phase space due to motion result in chaos-like behavior. We likewise demonstrate that the effects of spatio-temporally varying illumination produces phase space trajectories of the pixel amplitudes which are clearly non-chaotic. We review the research tying chaotic behavior to the fractal characteristics of phase space trajectories, and we investigate multi-fractal measures which can be used to classify the pixels in an image stream based on their fractal behavior in phase space. We finally apply these measures to the task of motion detection and segmentation and show they are effective in identifying moving objects while ignoring spatio-temporally varying illumination changes.

Embedded Active Vision System Based on an FPGA Architecture

In computer vision and more particularly in vision processing, the impressive evolution of algorithms and the emergence of new techniques dramatically increase algorithm complexity. In this paper, a novel FPGA-based architecture dedicated to active vision (and more precisely early vision) is proposed. Active vision appears as an alternative approach to deal with artificial vision problems. The central idea is to take into account the perceptual aspects of visual tasks, inspired by biological vision systems. For this reason, we propose an original approach based on a system on programmable chip implemented in an FPGA connected to a CMOS imager and an inertial set. With such a structure based on reprogrammable devices, this system admits a high degree of versatility and allows the implementation of parallel image processing algorithms.

Invariance and selectivity in the ventral visual pathway

Pattern recognition systems that are invariant to shape, pose, lighting and texture are never sufficiently selective; they suffer a high rate of “false alarms”. How are biological vision systems both invariant and selective? Specifically, how are proper arrangements of sub-patterns distinguished from the chance arrangements that defeat selectivity in artificial systems? The answer may lie in the nonlinear dynamics that characterize complex and other invariant cell types: these cells are temporarily more receptive to some inputs than to others ( functional connectivity). One consequence is that pairs of such cells with overlapping receptive fields will possess a related property that might be termed functional common input. Functional common input would induce high correlation exactly when there is a match in the sub-patterns appearing in the overlapping receptive fields. These correlations, possibly expressed as a partial and highly local synchrony, would preserve the selectivity otherwise lost to invariance.

NMF, LNMF, and DNMF modeling of neural receptive fields involved in human facial expression perception

Recently, three learning algorithms, namely non-negative matrix factorization (NMF), local non-negative matrix factorization (LNMF), and discriminant non-negative matrix factorization (DNMF) have been proposed to produce sparse image representations. However, when their input is a database of human facial images, they decompose the images into sparse representations with quite different degree of sparseness. Within a continuum of sparseness ranging from holistic to local image representation, the first algorithm rather tends towards the first extreme, while the second algorithm produces a local representation. The third algorithm provides an image representation that is in between these two extremes. These algorithms decompose the facial images in the database into basis images and their corresponding coefficients. The basis images are learned by the algorithm when human face images are given as input. By analogy to neurophysiology, the basis images could be associated with the receptive fields of neuronal cells involved in encoding human faces. Taken from this point of view, the paper presents an analysis of these three representations in connection to the receptive field parameters such as spatial frequency, frequency orientation, position, length, width, aspect ratio, etc. By analyzing the tiling properties of these bases we can have an insight of how suitable these algorithms are to resemble biological visual perception systems.

Lightness and Perceptual Transparency

To estimate intrinsic descriptors of objects in the environment, effective biological vision systems must 'discount' extrinsic image properties that arise from changes in viewing conditions. In particular, to estimate the reflectance of surfaces, human vision must discount, or 'take account of', likely differences in the illumination of surfaces between one image region and another. If human vision possesses any significant degree of lightness constancy, then we would expect a target perceived to be in low illumination to appear lighter than an identical target perceived to be in higher illumination. In this paper, I present lightness illusions that run directly counter to this expectation. I suggest that mid-level and higher-level factors such as image junction structure and perceived illumination and transparency, are ineffective for generating strong lightness illusions on their own, and that these factors are not 'stronger' than luminance contrast in determining lightness. I discuss the implications of these results for current models of lightness perception. I also suggest a statistical justification for the highest-luminance anchoring rule for lightness.

Topographic Product Models Applied to Natural Scene Statistics

We present an energy-based model that uses a product of generalized Student-t distributions to capture the statistical structure in data sets. This model is inspired by and particularly applicable to "natural" data sets such as images. We begin by providing the mathematical framework, where we discuss complete and overcomplete models and provide algorithms for training these models from data. Using patches of natural scenes, we demonstrate that our approach represents a viable alternative to independent component analysis as an interpretive model of biological visual systems. Although the two approaches are similar in flavor, there are also important differences, particularly when the representations are overcomplete. By constraining the interactions within our model, we are also able to study the topographic organization of Gabor-like receptive fields that our model learns. Finally, we discuss the relation of our new approach to previous work--in particular, gaussian scale mixture models and variants of independent components analysis.

Regularization operators for natural images based on nonlinear perception models

Image restoration requires some a priori knowledge of the solution. Some of the conventional regularization techniques are based on the estimation of the power spectrum density. Simple statistical models for spectral estimation just take into account second-order relations between the pixels of the image. However, natural images exhibit additional features, such as particular relationships between local Fourier or wavelet transform coefficients. Biological visual systems have evolved to capture these relations. We propose the use of this biological behavior to build regularization operators as an alternative to simple statistical models. The results suggest that if the penalty operator takes these additional features in natural images into account, it will be more robust and the choice of the regularization parameter is less critical.

Neuronal Mechanisms for Illusory Brightness Perception in Humans

Biological visual systems are extraordinarily capable of recovering the shape and brightness of objects from sparse and fragmentary information. Using functional magnetic imaging, we show that two associative areas of the dorsal pathway--in the caudal region of the intrapariatal sulcus and in the lateral occipital sulcus--respond specifically to the Craik-O'Brien-Cornsweet illusion generated by high-pass filtered edges. Other visual areas, including primary visual cortex, also respond strongly to the retinotopic location of the edge, but these areas respond equally well to a line of matched contrast and detectability, rather than specifically to the brightness illusion. The reconstruction of surface and/or its brightness seems to be achieved by associative areas from the information about visual features provided by the primary visual cortices, even where there is no physical difference in luminance.

Higher-level mechanisms detect facial symmetry

The role of symmetry detection in early visual processing and the sensitivity of biological visual systems to symmetry across a wide range of organisms suggest that symmetry can be detected by low-level visual mechanisms. However, computational and functional considerations suggest that higher-level mechanisms may also play a role in facial symmetry detection. We tested this hypothesis by examining whether symmetry detection is better for faces than comparable patterns, which share low-level properties with faces. Symmetry detection was better for upright faces than for inverted faces (experiment 1) and contrast-reversed faces (experiment 2), implicating high-level mechanisms in facial symmetry detection. In addition, facial symmetry detection was sensitive to spatial scale, unlike low-level symmetry detection mechanisms (experiment 3), and showed greater sensitivity to a 45 degrees deviation from vertical than is found for other aspects of face perception (experiment 4). These results implicate specialized, higher-level mechanisms in the detection of facial symmetry. This specialization may reflect perceptual learning resulting from extensive experience detecting symmetry in faces or evolutionary selection pressures associated with the important role of facial symmetry in mate choice and 'mind-reading' or both.

Geometric Fourier Analysis for Computational Vision

Projective Fourier analysis — geometric Fourier analysis of the group SL(2,ℂ), the group identified in the conformal camera that provides image perspective transformations—is discussed in the framework of representation theory of semisimple Lie groups. The compact model of projective Fourier analysis is constructed, complementing the noncompact model proposed before. Detailed mathematical formulation of both models is presented. It is demonstrated that the projective Fourier analysis provides the data model for efficient perspectively covariant digital image representation well adapted to the retino-cortical mapping of biological visual system, and therefore, explicitly designed for the foveated sensors of a silicon retina, the use of which in active vision systems is presently limited due to the lack of such a model.

ハイブリッド画像理解システムにおける選択的注意モデルのパラメータチューニング

Fukushima's Selective Attention Model is a model of biological vision system. It has engineering merit of deformation and position shift tolerant recognition and recollection of the recognized object in an input image. Based on this feature, we have proposed a method for hybrid image understanding in which each object is recognized, recollected and segmented sequentially even when the objects are overlapped and occluded. However, parameter setting for the fine object recollection is difficult with the Selective Attention Model and parameters search is necessary to tune the given image. In this paper, we propose a method for the parameter search based on an evaluation of image recollection precision in the hybrid image understanding process. A computer simulation result demonstrates the validity of the proposed method.

A Network of Analog Metal Oxide Semiconductor Circuits for Motion Detection with Local Adaptation to a Background Image

Inspired by a mechanism of biological vision systems, a model and a network of analog metal oxide semiconductor (MOS) circuits are proposed which display an optical flow with local adaptation to the relative velocity of a background image. A function of displaying an optical flow successfully worked as a result of simulations using the simulation program with integrated circuit emphasis (SPICE). A function of varying an optical flow at a certain instant by local adaptation was also demonstrated. The proposed network is suitable for the realization of a large-scale integrated circuit (LSI), which displays an optical flow with local adaptation to the local velocity of a background.

Hebbian encoding in the biological visual system.

We examined neural networks built of several hundred Hodgkin-Huxley neurons. The main aim of the research described below was to simulate memory processes occurring in hippocampus and biological visual system. In our model we chose the ancient Chinese I-Ching Oracle as a set of input patterns. Maps of Hebbian weights appearing on the output device of the model can be analysed by artificial neural networks playing a role of some kind of visual consciousness.

VLSI Implementation of a Neuromorphic Spiking Pixel and Investigation of Various Focal-Plane Excitation Schemes

In this paper we describe a Neuromorphic spiking pixel architecture allowing the conversion of the light intensity into a pulse train signal. We first describe the spiking pixel architecture and its inherent advantages such as light adaptation mechanism and linear response characteristics. Inspired from biological visual systems and the integrate and fire oscillator, different inter-pixels interaction schemes are studied and their synchronization ability are compared. The spiking pixel architecture was implemented and the functionality of the spiking pixel imager is demonstrated through experimental results.

Inertial Sensed Ego‐motion for 3D Vision

AbstractInertial sensors attached to a camera can provide valuable data about camera pose and movement. In biological vision systems, inertial cues provided by the vestibular system are fused with vision at an early processing stage. In this article we set a framework for the combination of these two sensing modalities. Cameras can be seen as ray direction measuring devices, and in the case of stereo vision, depth along the ray can also be computed. The ego‐motion can be sensed by the inertial sensors, but there are limitations determined by the sensor noise level. Keeping track of the vertical direction is required, so that gravity acceleration can be compensated for, and provides a valuable spatial reference. Results are shown of stereo depth map alignment using the vertical reference. The depth map points are mapped to a vertically aligned world frame of reference. In order to detect the ground plane, a histogram is performed for the different heights. Taking the ground plane as a reference plane for the acquired maps, the fusion of multiple maps reduces to a 2D translation and rotation problem. The dynamic inertial cues can be used as a first approximation for this transformation, allowing a fast depth map registration method. They also provide an image independent location of the image focus of expansion and center of rotation useful during visual based navigation tasks. © 2004 Wiley Periodicals, Inc.

Attentional sequence-based recognition: markovian and evidential reasoning

Biological vision systems explore their environment via allocating their visual resources to only the interesting parts of a scene. This is achieved by a selective attention mechanism that controls eye movements. The data thus generated is a sequence of subimages of different locations and thus a sequence of features extracted from those images - referred to as attentional sequence. In higher level visual processing leading to scene cognition, it is hypothesized that the information contained in attentional sequences are combined and utilized by special mechanisms - although still poorly understood. However, developing models of such mechanisms prove out to be crucial - if we are to understand and mimic this behavior in robotic systems. In this paper, we consider the recognition problem and present two approaches to using attentional sequences for recognition: Markovian and evidential reasoning. Experimental results with our mobile robot APES reveal that simple shapes can be modeled and recognized by these methods - using as few as ten fixations and very simple features. For more complex scenes, longer attentional sequences or more sophisticated features may be required for cognition.

A feature-based model of symmetry detection.

Symmetry detection is important for many biological visual systems, including those of mammals, insects and birds. We constructed a symmetry-detection algorithm with two stages: location of the visually salient features of the image, then evaluating the symmetry of these features over a long range, by means of a simple Gaussian filter. The algorithm detects the axis of maximum symmetry for human faces (or any arbitrary image) and calculates the magnitude of the asymmetry. We have evaluated the algorithm on the dataset of Rhodes et al. (1998 Psychonom. Bull. Rev. 5, 659-669) and found that the algorithm is able to discriminate small variations of symmetry created by computer-manipulating the symmetry levels in individual faces, and that the values measured by the algorithm correlate well with human psycho-physical symmetry ratings.

Bubbles: a unifying framework for low-level statistical properties of natural image sequences.

Recently, different models of the statistical structure of natural images have been proposed. These models predict properties of biological visual systems and can be used as priors in Bayesian inference. The fundamental model is independent component analysis, which can be estimated by maximization of the sparsenesses of linear filter outputs. This leads to the emergence of principal simple cell properties. Alternatively, simple cell properties are obtained by maximizing the temporal coherence in natural image sequences. Taking account of the basic dependencies of linear filter outputs permit modeling of complex cells and topographic organization as well. We propose a unifying framework for these statistical properties, based on the concept of spatiotemporal activity "bubbles."A bubble means here an activation of simple cells (linear filters) that is contiguous both in space (the cortical surface) and in time.

Surface color perception in constrained environments

AbstractByrne & Hilbert propose that color can be identified with explicit properties of physical surfaces. I argue that this claim must be qualified to take into account constraints needed to make recovery of surface color information possible. When these constraints are satisfied, then a biological visual system can establish a correspondence between perceived surface color and specific surface properties.

Materials for Bio-inspired Optics

Bio-inspired optical systems have the potential to transform the process of image collection and analysis in the future. Through simplicity and dynamics these collection systems may provide multiple functions with an excellent level of reliability. From the graded index of the fish-eye to the hyper-acuity of the compound eye, nature has provided function with amazing simplicity. This article reviews a select number of biological vision systems and the unique material behaviors that rise to these characteristics. We introduce the materials chemistry and morphological complexity that enables the biological system behavior.