Filter bank is the most common feature being employed in the research of the marginalisation approaches for robust speech recognition due to its simplicity in detecting the unreliable data in the frequency domain. In this paper, we propose a hybrid approach based on the marginalisation and the soft decision techniques that make use of the Mel-frequency cepstral coefficients (MFCCs) instead of f...

In this paper, a novel orthogonal 2-D lattice structure is incorporated into the design of a nonseparable 2-D fourchannel perfect reconstruction filter bank. The proposed filter bank is obtained by using the polyphase decomposition technique which requires the design of an orthogonal 2-D lattice filter. Due to constraint of perfect reconstruction, each stage of this lattice filter bank is simpl...

The mapping of complex-valued functions onto phase-only spatial light modulators is examined. Random phase encoding effectively adds amplitude control to the phase-only filter and can be used to trade off systematic errors of the phase-only filter for random errors. This is illustrated for the problem of recognizing a threedimensional object from arbitrary views. The complex-valued composite fi...

A versatile filter-bank concept for adaptive subband filtering is proposed, which achieves a significantly lower algorithmic signal delay than commonly used analysissynthesis filter-banks. It is derived as an efficient implementation of the filter-bank summation method and performs time-domain filtering with coefficients adapted in the uniform or non-uniform frequency-domain. The frequency warp...

A novel nonuniform cosine modulated filter bank is presented, mixing the concepts of warped DFT (Discrete Fourier Transform) filter bank and pseudoQMF system (Quadrature Mirror Filter). Unequal bandwidths are achieved by allpass transformation of uniform polyphase system. The cosine modulation results in real valued channel signals, and hence simplified subband processing. The considered filter...

Automatic recognition of human emotion has become an interesting topic among brain-computer interface (BCI) researchers. Emotion is one the most fundamental features a subject. With proper analysis emotion, inner state subject can be assessed directly. The brain response competently represented by electroencephalography (EEG). selection potential in EEG related to very important task for develo...

Recently local frequency has been suggested as an efficient feature for performing image analysis and multi-modal image registration. While local frequency representation can detect the structure of the scene in the image (ridges and edges simultaneously), its computation typically requires using a bank of analytic filters such as two-dimensional (2-D) Gabor filters characterized by a set of pa...

In this study, a novel regularized common spatial pattern method is introduced. Spatial filtering is an important processing step for feature extraction in motor imagery based brain computer interfaces. Common Spatial Patterns (CSP) method is an effective spatial filter for discriminating different motor imagery signals acquired using large number of EEG electrodes. Unfortunately, CSP method is...

This paper presents enhancement of image reconstruction by comprehensively exploiting the phase diversity of adaptive synthesis filter bank. A novel approach of selection of synthesis delay filters in various reconstruction groups and possible placing configurations of optimal groups in adaptive synthesis filter bank have been explored. In contrast with linear phase synthesis filters, synthesis...

We consider the use of pre and postfilters in conjunction with M-channel, uniform-hand paraunitary (orthonormal) filter banks. We show that given any orthonormal filter bank, the pre and postfilters that maximize the coding gain are determined entirely by the power spectrum of the input process regardless of the details of the orthainormal filter bank (which could be FIR, IIR, or even the ideal...