Document representation has a large impact on the performance of document retrieval and clustering algorithms. We propose a hybrid document indexing scheme that combines the traditional bagof-words representation with spectral embedding. This method accounts for the specifics of the document collection and also uses semantic similarity information based on a large scale statistical analysis. Cl...

We have investigated the use of a pitch adaptive spectral representation on large vocabulary speech recognition, in conjunction with speaker normalisation techniques. We have compared the effect of a smoothed spectrogram to the pitch adaptive spectral analysis by decoupling these two components of STRAIGHT. Experiments performed on a large vocabulary meeting speech recognition task highlight th...

Methods are developed for constructing spectral representations of cold (barotropic) neutron-star equations of state. These representations are faithful in the sense that every physical equation of state has a representation of this type and conversely every such representation satisfies the minimal thermodynamic stability criteria required of any physical equation of state. These spectral repr...

In [M. K. Vemuri, Realizations of the canonical representation, submitted], it was shown that the spectral synthesis problem for the Alpha transform is closely related to the problem of classifying realizations of the canonical representation (of the Heisenberg group). In this paper, we show that discrete sets are sets of spectral synthesis for the Alpha transform.

The purpose of the research which led to this paper is to develop an analytical procedure for the design of the spectral channels for multispectral remote sensor systems. An optimum design based on the criterion of minimum meansquare representation error using the KarhunenLoeve expansion was developed to represent the spectral response functions from a stratum. From the overall pattern recognit...

The central representation of a given acoustic motif is thought to be strongly context dependent, i.e., to rely on the spectrotemporal past and present of the acoustic mixture in which it is embedded. The present study investigated the cortical representation of spectral edges (i.e., where stimulus energy changes abruptly over frequency) and its dependence on stimulus duration and depth of the ...

A type of representation of the spectral bidirectional reflectance distribution function (BRDF) is proposed that distinctly separates the spectral variable (wavelength) from the geometrical variables (spherical coordinates of the irradiation and viewing directions). Principal components analysis (PCA) is used in order to decompose the spectral BRDF in decorrelated spectral components, and the w...