E.D. Di Claudio

Car plate recognition by neural networks and image processing

In this paper we describe an experimental system for the recognition of Italian-style car license plates. Images are usually taken from a camera at a toll gate and preprocessed by a fast and robust 1-D DFT scheme to find the plate and character positions. Characters are classified by a multilayer neural network trained by the recently developed BRLS learning algorithm. The same neural network replaces both the traditional feature extractor and the classifier. The percentage of correctly recognized characters reaches the best scores obtained in literature, being highly insensitive to the environment variability, while the architecture appears best suited for parallel implementation on programmable DSP processors.

A generalized learning paradigm exploiting the structure of feedforward neural networks

In this paper a general class of fast learning algorithms for feedforward neural networks is introduced and described. The approach exploits the separability of each layer into linear and nonlinear blocks and consists of two steps. The first step is the descent of the error functional in the space of the outputs of the linear blocks (descent in the neuron space), which can be performed using any preferred optimization strategy. In the second step, each linear block is optimized separately by using a least squares (LS) criterion. To demonstrate the effectiveness of the new approach, a detailed treatment of a gradient descent in the neuron space is conducted. The main properties of this approach are the higher speed of convergence with respect to methods that employ an ordinary gradient descent in the weight space backpropagation (BP), better numerical conditioning, and lower computational cost compared to techniques based on the Hessian matrix. The numerical stability is assured by the use of robust LS linear system solvers, operating directly on the input data of each layer. Experimental results obtained in three problems are described, which confirm the effectiveness of the new method.

Fast adaptive digital equalization by recurrent neural networks

Neural networks (NNs) have been extensively applied to many signal processing problems. In particular, due to their capacity to form complex decision regions, NNs have been successfully used in adaptive equalization of digital communication channels. The mean square error (MSE) criterion, which is usually adopted in neural learning, is not directly related to the minimization of the classification error, i.e., bit error rate (BER), which is of interest in channel equalization. Moreover, common gradient-based learning techniques are often characterized by slow speed of convergence and numerical ill conditioning. In this paper, we introduce a novel approach to learning in recurrent neural networks (RNNs) that exploits the principle of discriminative learning, minimizing an error functional that is a direct measure of the classification error. The proposed method extends to RNNs a technique applied with success to fast learning of feedforward NNs and is based on the descent of the error functional in the space of the linear combinations of the neurons (the neuron space); its main features are higher speed of convergence and better numerical conditioning w.r.t. gradient-based approaches, whereas numerical stability is assured by the use of robust least squares solvers. Experiments regarding the equalization of PAM signals in different transmission channels are described, which demonstrate the effectiveness of the proposed approach.

Multi-source localization in reverberant environments by ROOT-MUSIC and clustering

Localization of acoustic sources in reverberant environments by microphone arrays remains a challenging task in audio signal processing. As a matter of fact, most assumptions of commonly adopted models are not met in real applications. Moreover, in practical systems it is not convenient or possible to employ sophisticated and costly architectures, that require precise synchronization and fast data shuffling among sensors. In this paper, a new robust multi-step procedure for speaker localization in reverberant rooms is introduced and described. The new approach is based on a disturbed harmonics model of time delays in the frequency domain and employs the well-known ROOT-MUSIC algorithm, after a preliminary distributed processing of the received signals. Candidate source positions are then estimated by clustering of raw TDOA estimates. Main features of the proposed approach, compared to previous solutions, are the capability of tracking multiple speakers and the high accuracy of the closed form TDOA estimator.

Asymptotically perfect wideband focusing of multiring circular arrays

The concept of coherent focusing of sensor arrays, introduced by Wang and Kaveh, led to the development of high-performance and computationally efficient algorithms for wideband direction finding and beamforming. Nonetheless, the quality of focusing depends on the understanding and the proper exploitation of specific array manifold properties. Circular arrays exhibit uniform (isotropic) performance over the entire azimuthal range and allow the use of fast algorithms, originally developed for uniform linear arrays, by decomposing the manifold into circular harmonics. Coherent wideband focusing of circular arrays suffers from ambiguities, noise warping, and numerical ill conditioning. In this work, it is shown that a fast orthonormal beamspace transformation combining the responses of two concentric rings can perfectly focus wideband sources at all azimuth angles in the circular harmonic domain in the limit of infinite number of sensors. The information loss after focusing is minimized through an analytical approach. The proposed focusing scheme is computationally very efficient and can be directly extended to multiring circular arrays that are arranged into a set of nested two-ring subarrays to cover very large bandwidths.

Robust ML wideband beamformingin reverberant fields

Adaptive beamforming of sensor arrays immersed into reverberant fields can easily result in the cancellation of the useful signal because of the temporal correlation existing among the direct and the reflected path signals. Wideband beamforming can somewhat mitigate this phenomenon, but adaptive solutions based on the minimum variance (MV) criterion remain nonrobust in many practical applications, such as multimedia systems, underwater acoustics, and seismic prospecting. In this paper, a steered wideband adaptive beamformer, optimized by a novel concentrated maximum likelihood (ML) criterion in the frequency domain, is presented and discussed in the light of a very general reverberation model. It is shown that ML beamforming can alleviate the typical cancellation problems encountered by adaptive MV beamforming and preserve the intelligibility of a wideband and colored source signal under interference, reverberation, and propagation mismatches. The difficult optimization of the ML cost function, which incorporates a robustness constraint to prevent signal cancellation, is recast as an iterative least squares problem through the concept of descent in the neuron space, which was originally developed for the training of multilayer neural networks. Finally, experiments with computer-generated and real-world data demonstrate the superior performance of the proposed beamformer with respect to its MV counterpart.

Two-Dimensional Approach to Full-Reference Image Quality Assessment Based on Positional Structural Information

A method for full-reference visual quality assessment based on the 2-D combination of two diverse metrics is described. The first metric is a measure of structural information loss based on the Fisher information about the position of the structures in the observed images. The second metric acts as a categorical indicator of the type of distortion that images underwent. These two metrics constitute the inner state of a virtual cognitive model, viewed as a system whose output is the automatic visual quality estimate. The use of a 2-D metric fills the intrinsic incompleteness of methods based on a single metric while providing consistent response across different image impairment factors and blind distortion classification capability with a modest computational overhead. The high accuracy and robustness of the method are demonstrated through cross-validation experiments.

Fast SVD-based algorithm for signal selective DOA estimation

The classical formulation of subspace direction of arrival (DOA) estimation algorithms, such as MUSIC, assumes temporally independent source signals and white noise background. If only a subset of source signals are of interest (SOI), a computationally expensive classification procedure should be performed after both DOA parameter estimation and signal extraction steps. In this paper, a simple and robust SVD-based estimation procedure is proposed, making use of an orthogonal two-channel decomposition of sensor signals in the time domain, optimized for the SOI. Simulation results are provided to support the capability of the proposed algorithm of discarding temporally uncorrelated interference and identifying the SOI DOA parameters only, without any need of post-processing.

Optimal weighted LS AR estimation in presence of impulsive noise

A procedure for assigning optimal weights to the prediction equations which are used to obtain the parameters of an autoregressive (AR) model for spectrum estimation by the least squares (LS) solution is presented. The set of weights is computed, by linear programming techniques, in order to reduce the effects of strong impulsive noise onto the AR parameter estimate. The method is particularly effective when the Gaussian white noise component is much smaller than both spikes and useful signal. In order to demonstrate the capability of the proposed approach, the results of a simple AR parameter estimation experiment are also reported.<<ETX>>

A clustering approach to multi-source localization in reverberant rooms

Localization of acoustic sources in the presence of reverberation is still a challenging task in audio signal processing. As a matter of fact, commonly adopted models are not adequate to describe real scenarios. Moreover, practical systems should not employ sophisticated and expensive architectures, that require precise synchronization and fast data shuffling among sensors. This work describes a new robust multi-step procedure for speaker localization in reverberant rooms. The proposed approach is based on a disturbed harmonics model of time delays in the frequency domain and employs the well-known ROOT-MUSIC algorithm, after a proper pre-processing of the received signals. Final clustering of raw TDOA estimates gives candidate source positions. Among the appealing features of the proposed approach are the capability of tracking multiple speakers simultaneously and the high accuracy of the closed form TDOA estimator.