Jeanny Hérault

Blind separation of sources, Part 1: an adaptive algorithm based on neuromimetic architecture

Abstract The separation of independent sources from an array of sensors is a classical but difficult problem in signal processing. Based on some biological observations, an adaptive algorithm is proposed to separate simultaneously all the unknown independent sources. The adaptive rule, which constitutes an independence test using non-linear functions, is the main original point of this blind identification procedure. Moreover, a new concept, that of INdependent Components Analysis (INCA), more powerful than the classical Principal Components Analysis (in decision tasks) emerges from this work.

Linear demosaicing inspired by the human visual system

There is an analogy between single-chip color cameras and the human visual system in that these two systems acquire only one limited wavelength sensitivity band per spatial location. We have exploited this analogy, defining a model that characterizes a one-color per spatial position image as a coding into luminance and chrominance of the corresponding three colors per spatial position image. Luminance is defined with full spatial resolution while chrominance contains subsampled opponent colors. Moreover, luminance and chrominance follow a particular arrangement in the Fourier domain, allowing for demosaicing by spatial frequency filtering. This model shows that visual artifacts after demosaicing are due to aliasing between luminance and chrominance and could be solved using a preprocessing filter. This approach also gives new insights for the representation of single-color per spatial location images and enables formal and controllable procedures to design demosaicing algorithms that perform well compared to concurrent approaches, as demonstrated by experiments.

Using Human Visual System modeling for bio-inspired low level image processing

An efficient modeling of the processing occurring at retina level and in the V1 visual cortex has been proposed in [1,2]. The aim of the paper is to show the advantages of using such a modeling in order to develop efficient and fast bio-inspired modules for low level image processing. At the retina level, a spatio-temporal filtering ensures accurate structuring of video data (noise and illumination variation removal, static and dynamic contour enhancement). In the V1 cortex, a frequency and orientation based analysis is performed. The combined use of retina and V1 cortex modeling allows the development of low level image processing modules for contour enhancement, for moving contour extraction, for motion analysis and for motion event detection. Each module is described and its performances are evaluated. The retina model has been integrated into a real-time C/C++ optimized program which is also presented in this paper with the derived computer vision tools.

An efficient neuromorphic analog network for motion estimation

Optical flow estimation is a critical mechanism for autonomous mobile robots as it provides a range of useful information. As real-time processing is mandatory in this case, an efficient solution is the use of specific very large scale integration (VLSI) analog circuits. This paper presents a simple and regular architecture based on analog circuits, which implements the entire processing line from photoreceptor to accurate and reliable optical flow estimation. The algorithm we propose, is an energy-based method using a novel wideband velocity-tuned filter which proves to be an efficient alternative to the well-known Gabor filters. Our approach shows that a high level of accuracy can be obtained from a small number of loosely tuned filters. It exhibits similar or improved performance to that of other existing algorithms, but with a much lower complexity.

A model of colour processing in the retina of vertebrates: From photoreceptors to colour opposition and colour constancy phenomena

Abstract Starting with a biological model of the foveal and parafoveal regions of the retina, this paper shows that simple considerations about the spatiotemporal filtering processed by the retinal neural network can account for colour processing at the level of early vision. We establish that the Red, Green, Blue signal can be considered as a low-pass luminance signal plus a colour-modulated signal. The structure of the Outer- and Inner Plexiform Layers of the retina leads to spatial low- and high-pass filters which account for the achromatic and transient characteristics of the Y ganglion cells as well as for the spatiotemporal colour-opponent properties of X ganglion cells. Considering this property and the logarithmic transduction of photoreceptors, it is easy to postulate that, after the retina at the level of the Lateral Geniculate Nucleus, a simple low-pass filtering can pave the way to the well known colour constancy phenomenon.

Analog VLSI synaptic matrices as building blocks for neural networks

An associative memory circuit that may let designers expand neural networks around a matrix of analog synapses is described. The architecture of the chip and its basic cell are discussed, and some SPICE simulation results are presented and compared with measures provided by the first prototype. In particular, the linearity and dynamic response of the complete chip, which includes an array of 25 synapses and two address decoders used for programming the weights, are examined.<<ETX>>

Modeling visual perception for image processing

This paper presents a model of the retina with its properties with respect to sampling, spatiotemporal filtering, color-coding and non-linearity, and their consequences on the processing of visual information. Its formalism points out the architectural and algorithmic principles of neuromorphic circuits which are known to improve compactness, consumption, robustness and efficiency, leading to direct applications in engineering science. Its biological aspect, strongly based neural and cellular descriptions makes it suitable as an investigation tool for neurobiologists, allowing the simulation of experiences difficult to set up and answering fundamental theoretical questions.

Curvilinear component analysis for high-dimensional data representation: I. Theoretical aspects and practical use in the presence of noise

Starting from a recall of the theoretical framework, this paper presents the conditions and the strategy of implementation of CCA, a recent algorithm for non-linear mapping. Initially developed in a basic form, for non-linear and high-dimensional data sets, the algorithm is here adapted to the general, and more realistic, case of noisy data. This algorithm, which finds the manifold (in particular, the intrinsic dimension) of the data, has proved to be very efficient in the representation of highly folded data structures. We describe here how it can be tuned to find the average manifold and how robust the convergence is. A companion paper (this issue) presents various applications using this property.

Spatio-temporal Tone Mapping Operator Based on a Retina Model

From moonlight to bright sun shine, real world visual scenes contain a very wide range of luminance; they are said to be High Dynamic Range (HDR). Our visual system is well adapted to explore and analyze such a variable visual content. It is now possible to acquire such HDR contents with digital cameras; however it is not possible to render them all on standard displays, which have only Low Dynamic Range (LDR) capabilities. This rendering usually generates bad exposure or loss of information. It is necessary to develop locally adaptive Tone Mapping Operators (TMO) to compress a HDR content to a LDR one and keep as much information as possible. The human retina is known to perform such a task to overcome the limited range of values which can be coded by neurons. The purpose of this paper is to present a TMO inspired from the retina properties. The presented biological model allows reliable dynamic range compression with natural color constancy properties. Moreover, its non-separable spatio-temporal filter enhances HDR video content processing with an added temporal constancy.

Practical implementation of LMMSE demosaicing using luminance and chrominance spaces

Most digital color cameras sample only one color at each spatial location, using a single sensor coupled with a color filter array (CFA). An interpolation step called demosaicing (or demosaicking) is required for rendering a color image from the acquired CFA image. Already proposed linear minimum mean square error (LMMSE) demosaicing provides a good tradeoff between quality and computational cost for embedded systems. In this paper we propose a modification of the stacked notation of superpixels, which allows an effective computing of the LMMSE solution from an image database. Moreover, this formalism is used to decompose the CFA sampling into a sum of a luminance estimator and a chrominance projector. This decomposition allows interpreting estimated filters in term of their spatial and chromatic properties and results in a solution with lower computational complexity than other LMMSE approaches for the same quality.