Gaetano Scarano

Discrete time techniques for time delay estimation

Basic aspects of time delay estimation (TDE) based on sampled signals are investigated. The direct cross-correlation method is analyzed and compared to the average square difference function (ASDF) and the (addition only based) average magnitude difference function (AMDF) estimators, Their relative accuracy is theoretically evaluated, and previous empirical results are explained. It is shown that both the ASDF- and the AMDF-based estimators outperform the direct cross-correlation based estimator for medium-high signal-to-noise ratios. Moreover, the AMDF-based estimator, which avoids any multiplications, significantly reduces the computational complexity of the estimation procedure while offering only a moderate performance loss with respect to the ASDF based estimator. >

Multichannel blind image deconvolution using the Bussgang algorithm: spatial and multiresolution approaches

This work extends the Bussgang blind equalization algorithm to the multichannel case with application to image deconvolution problems. We address the restoration of images with poor spatial correlation as well as strongly correlated (natural) images. The spatial nonlinearity employed in the final estimation step of the Bussgang algorithm is developed according to the minimum mean square error criterion in the case of spatially uncorrelated images. For spatially correlated images, the nonlinearity design is rather conducted using a particular wavelet decomposition that, detecting lines, edges, and higher order structures, carries out a task analogous to those of the (preattentive) stage of the human visual system. Experimental results pertaining to restoration of motion blurred text images, out-of-focus spiky images, and blurred natural images are reported.

A multiresolution approach for texture synthesis using the circular harmonic functions

In this paper, an unsupervised model-based texture reproduction technique is described. In accordance with the Juleszs (1962) conjecture, the statistical properties of the prototype up to the second order are copied in order to generate a synthetic texture perceptually indistinguishable from the given sample. However, this task is accomplished using a hybrid approach which operates partially in the spatial domain and partially in a multiresolution domain. The latter employed is the circular harmonic function (CHF) domain since it has been proven to be well suited for mimicking the behavior of the human visual system (HVS). This approach allows, for a wide range of textures typologies, obtaining synthetic textures that better match the prototype with respect to the ones obtained using techniques based on the Juleszs conjecture operating only in the spatial domain, and to dramatically reduce the computational complexity of similar methods operating only in the multiresolution domain.

Texture synthesis-by-analysis with hard-limited Gaussian processes

A twin stage texture synthesis-by-analysis method is presented. It aims to approximate first- and second-order distributions of the texture, accordingly to the Julesz conjecture. In the first stage, the binary textural behavior of a given prototype is represented by means of a hard-limited Gaussian process. In the second stage, the texture is synthesized by passing the binary hard-limited Gaussian process through a linear filter followed by a zero memory histogram equalizer.

Robust rotation-invariant texture classification using a model based approach

In this paper, a model based texture classification procedure is presented. The texture is modeled as the output of a linear system driven by a binary image. This latter retains the morphological characteristics of the texture and it is specified by its spatial autocorrelation function (ACF). We show that features extracted from the ACF of the binary excitation suffice to represent the texture for classification purposes. Specifically, we employ a moment invariants based technique to classify the ACF. The resulting proposed classification procedure is thus inherently rotation invariant. Moreover, it is robust with respect to additive noise. Experimental results show that this approach allows obtaining high correct rotation-invariant classification rates while containing the size of the feature space.

Cumulant series expansion of hybrid nonlinear moments of complex random variables

A general theorem for zero-memory nonlinear transformations of complex stochastic processes is presented. It is shown that, under general conditions, the cross covariance between a stochastic process and a distorted version of another process can be represented by a series of cumulants. The coefficients of this cumulant expansion are expressed by the expected values of the partial derivatives, appropriately defined, of the function describing the nonlinearity. The theorem includes as a particular case the invariance property (Bussgangs (1952) theorem) of Gaussian processes, while holding for any joint distribution of the processes. The expansion in cumulants constitutes an effective means of analysis for higher-order-moment-based estimation procedures involving non-Gaussian complex processes. >

Brain waves based user recognition using the “eyes closed resting conditions” protocol

In this paper the use of brain waves as a biometric identifier is investigated. Among the very different protocols that can be used to acquire the electroencephalogram signal (EEG) of an individual we rely on a very simple one: closed eyes in resting conditions. A database of 48 healthy subjects, collected by the authors at the neurophysiology laboratory of the IRCCS Fondazione Santa Lucia, Roma, Italy, has been used for the experiments. Signals acquired from triplets of electrodes have been employed in the experimentations. In more detail, ten different triplets have been used separately in the experiments in order to speculate about the most suitable triplet to capture the occurring phenomena. Feature vectors constituted by the reflection coefficients of a six order AR model have been extracted for each used channel thus giving rise to a feature vector of length eighteen. A polynomial regression based classification is then employed. This analysis has been performed for three different frequency bands for each of the ten different triplet under analysis. The obtained genuine acceptance rate is of 96.08%.

Brain activity during the memorization of visual scenes from TV commercials: an application of high resolution EEG and steady state somatosensory evoked potentials technologies.

The aim of this study was to elucidate if the TV commercials that were remembered by the subjects after their observation within a documentary elicited particular brain activity when compared to the activity generated during the observation of TV commercials that were forgotten. High resolution EEG recordings were performed in a group of 10 healthy subjects with the steady state somatosensory evoked potentials (SSSEPs) technique, in which a series of light electrical stimulation at the left wrist were delivered at the frequency of 20Hz. The brain activity was indexed by the phase delay of the EEG spectral responses at 20Hz with respect to the stimulus delivering and evaluated at the scalp level as well as at the cortical surface using several regions of interest coincident with the Brodmann areas (BAs). Results suggest that the cerebral processes involved during the observation of TV commercials that were remembered by the population examined (RMB dataset) are generated by the posterior parietal cortices and the prefrontal areas, rather bilaterally. These results are compatible with previously results obtained in literature by using MEG and fMRI devices during similar experimental tasks. High resolution EEG is able to summarize, with the use of SSSEPs methodologies, the behavior of the estimated cortical networks subserving the proposed memory tasks. It is likely that such tool could play a role in the next future for the investigation of the neural substrates of the human behavior in decision-making and recognition tasks.

Blind phase recovery for QAM communication systems

In this paper, a novel phase estimator that can be employed for both square and cross Quadrature Amplitude Modulation (QAM) based digital transmission is presented. It does not need gain control and requires only the knowledge of the type of the transmitted symbol constellation, i.e., square or cross. It is based on the evaluation of the fourth power of the received data and the measurement of the orientation of the concentration ellipses of the bivariate Gaussian distribution having the same second-order moments. The analytical evaluation of the estimation error as well as of the asymptotic variance is provided. Experimental results outline the good performance of the estimator described here, which is superior to that of well-known phase estimation methods. Finally, it is outlined how the eccentricity of the concentration ellipses can be used to devise a test for detecting the constellation type.

Transient signal detection using higher order moments

The asymptotic performance of transient detection based on higher order moments is theoretically derived, invoking the asymptotic normality of the decision statistics, and a detector based on the third-order absolute moment is proposed. The analysis shows that for a transient duration of less than half the observation window, the proposed detector outperforms the detectors based on the second- and fourth-order moments for a wide range of SNR values. Computer simulations assess the applicability of the theoretical analysis.