The Infona portal uses cookies, i.e. strings of text saved by a browser on the user's device. The portal can access those files and use them to remember the user's data, such as their chosen settings (screen view, interface language, etc.), or their login data. By using the Infona portal the user accepts automatic saving and using this information for portal operation purposes. More information on the subject can be found in the Privacy Policy and Terms of Service. By closing this window the user confirms that they have read the information on cookie usage, and they accept the privacy policy and the way cookies are used by the portal. You can change the cookie settings in your browser.
A method is proposed for creating dictionaries of near-perfect reconstruction wavelets. Choosing a wavelet from a wavelet dictionary is an attractive alternative to customizing a wavelet for a specific task. The method is based on a modified simulated annealing scheme that implements unconstrained minimization. The amount of vanishing moments for the decomposition wavelets was fixed before the iteration...
In this paper, we improve on the slow time-varying partial correlation (STV-PARCOR) model recently suggested by us [2] to include any deterministic interpolator. We then suggest a modification to the on-line filtering algorithm to accomodate the changes. It is believed that the modification will improve on the simulation results as it takes into account the underlying trend of the parameter evolution...
We propose a new scheme for adapting support vector classifiers (SVC) in non-stationary environments. This adaptive SVC (ASVC) relies in interpreting the margin and penalty factor of the SVC as a relevance measure over the samples and on an iterative re-weighted least square (IRWLS) approach for optimizing it, which resembles the RLS filtering for adaptive equalization. The ASVC capabilities are shown...
Digital watermarking schemes embed additional information into digital data and are used in various applications, such as proof of ownership or fingerprinting. For such applications, the presence of watermarks must be provable to any possibly dishonest party. Standard watermark detection requires knowledge of sensitive information like the watermark or the embedding key. This is a major security risk,...
We develop two approaches for performing accurate traveltime in a 2D lattice. These methods are extensions of the Fast Marching method and consist in considering the 8-connected neighbors to compute the solution of the Eikonal equation. The interest of our approach is demonstrated by comparing the results obtained with the classical Fast Marching algorithm and with our methods in the case of a simple...
In this paper, we present a new fast algorithm for two-dimensional (2-D) linear adaptive filtering using the fast Chandrasekhar equations. Using the analogy between the multichannel linear model and the 2-D one, we transform an image into multichannel sequence and we extend the fast Chandrasekhar adaptive multichannel filtering algorithm to the 2-D case i.e. image filtering. The performance of the...
When the implementation of an acoustic echo canceler is implemented in the Mobile Switching Center of a GSM network, it is necessary to take into account the non linearities introduced by the speech coder/decoder. A classical linear adaptive filter is not sufficient. The performances of a classical AEC are improved by noise reduction adaptive post-filter. The post-filter introduced, take into account...
Multichannel audio offers significant advantages for music reproduction that include the ability to provide better localization and envelopment, as well as reduced imaging distortion. On the other hand, multichannel audio is one of the most demanding media types in terms of transmission requirements. A novel architecture was previously proposed, allowing delivery of uncompressed multichannel audio...
We address the problem of matched subspace detection in the presence of arbitrary noise and interferents, or interfering signals that may lie in a possibly unknown subspace, but that nevertheless corrupt the measurements. A hypothesis test that is robust to interferents yet sensitive to the signal of interest is formulated. The test is applicable to a large class of noise density functions. In addition,...
In this paper we propose a novel approach to sound reverberation based on the the geometric analysis of the acoustic environment, which allows the listener to freely move within it. The method is based on a combination of a tapped delay line for early reverberation, designed through beam tracing; and a wave digital network for late reverberation, designed through path tracing. The method is efficient...
The bispectrum is crucial for description of non-Gausssian and/or non-linear signals. In this paper we propose wavelet-thresholding estimators of the bispectrum of zero-mean, non-Gaussian, stationary signals. It is known in the case of Gaussian regression that wavelet estimators outperform traditional linear methods if the regularity of the function to be estimated varies substantially over its domain...
The Weights-Adjusted Second-Order Blind Identification (WASOBI) algorithm was recently proposed (Yeredor, 2000) as an optimized version of the SOBI Algorithm (Belouchrani et al., 1997) for blind separation of static mixtures of Gaussian Moving Average (MA) sources. The optimization consists of transforming the approximate joint diagonalization in SOBI into a properly weighted Least-Squares problem,...
In this paper, we analyse the stability of LMS algorithms with delayed coefficient adaptation, such as FxLMS and DLMS algorithms, when sinusoidal references are used. We demonstrate that in two common situations, that are two in-phase and quadrature sinusoidal components and a tapped delay line with a large number of weights, the adaptive system turns to be linear and time-invariant. Without the need...
This paper deals with an entropic approach as unsupervised thresholding technique for image processing, in order to extract a relevant binary information from noisy data. The method is based on the computation of the entropy power of the information source, as defined by Shannon. The threshold used for binarization is proportional to the entropic deviation of the observation source. The performance...
An extension to connected operator filtering is proposed here for 3D still images (e.g. biomedical volumes). The “surface preservation” property of this kind of morphological processing can be exploited to obtain a “simplification” effect on the data, deriving from the cancellation of some undesired 3D structures; this is achieved by some filtering criteria which can be related to different objectives,...
The object of this work is the application of steerable filters for the diagnosis of breast cancer. Because of the very specific geometry of malignant densities, we intend to apply a multiscale and orientation shiftable method for the detection of breast cancer in mammograms. Stellate masses have an irregular appearance and are frequently surrounded by a radiating pattern of linear spicules. Multiscale...
This paper addresses the problem of noise estimation in the context of speech processing. A recently proposed quantile-based approach to noise estimation has the merit of not relying on the explicit detection of speech, non-speech boundaries. Here this approach is extended to both time and frequency. The resultant time-frequency quantile-based noise estimation is shown to give superior ASR performance...
Blind and semiblind antenna array algorithms are proposed for Orthogonal Frequency Division Multiplexing (OFDM) systems in the time domain. A blind technique is proposed for the maximization of the temporal Signal to Noise plus Interference Ratio (SNIR) in frequency-selective channels exploiting the Cyclic Prefix (CP). In OFDM systems, this maximization does not imply the minimization of the Bit Error...
This paper presents a procedure to improve the quality of narrowband (0–4khz) CELP coded speech. The procedure is based to refine the pitch periodicity and reinsert the high frequency components (4–8khz) in the narrowband CELP decoded speech. The narrowband CELP decoded speech is first analyzed with Harmonic+Noise analyzer and Lowband information are extracted. By exploiting the Lowband spectrum envelope...
Set the date range to filter the displayed results. You can set a starting date, ending date or both. You can enter the dates manually or choose them from the calendar.