In this paper we study the performance of the low-variance multi-taper Mel-frequency cepstral coefficient (MFCC) and perceptual linear prediction (PLP) features in a state-ofthe-art i-vector speaker verification system. The MFCC and PLP features are usually computed from a Hamming-windowed periodogram spectrum estimate. Such a singletapered spectrum estimate has large variance, which can be red...

for a fiber optical surface plasmon resonance (spr) sensor a short part of its cladding should be removed to coat a thin layer of a metal. usually this is problematic when an optical fiber with small core diameter is used. in this paper, a new method using µliter droplet of the hf acid for short fiber optical taper fabrication is reported. using this method in a multi-mode optical fiber with th...

This paper presents a method to recover task-related sources from a multi-class BrainComputer Interface (BCI) based on motor imagery. Our method gathers two common approaches to tackle the multi-class problem: 1) the supervised approach of Common Spatial Pattern (CSP) to discriminate between different tasks; 2) the criterion of statistical independence of non-stationary sources used in Independ...

We develop and characterize a UV ablation technique that can be used to pattern soft materials such as polymers and nonlinear molecules self-assembled over silica microstructures. Using this method, we fabricate a spatially periodic coating of nonlinear film over a thin silica fiber taper for second harmonic generation (SHG). Experimentally, we find that the second harmonic signal produced by t...

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...

Likelihood approaches for large irregularly spaced spatial datasets are often very difficult, if not infeasible, to use due to computational limitations. Even when we can assume normality, exact calculations of the likelihood for a Gaussian spatial process observed at n locations requires O(n) operations. We present a version of Whittle’s approximation to the Gaussian log likelihood for spatial...