High dimensional imagery geometry and it's applications

Because of information digitalization and the correspondence of digits and the coordinates, Information Science and high-dimensional space have consanguineous relations. With the transforming from the information issues to the point analysis in high-dimensional space, we proposed a novel computational theory, named High dimensional imagery geometry (HDIG). Some computational algorithms of HDIG have been realized using software, and how to combine with groups of simple operators in some 2D planes to implement the geometrical computations in high-dimensional space is demonstrated in this paper. As the applications, two kinds of experiments of HDIG, which are blurred image restoration and pattern recognition ones, are given, and the results are satisfying.

Research on continuous speech recognition with small vocabulary based on biomimetic pattern recognition

Based on biomimetic pattern recognition theory, we proposed a novel speaker-independent continuous speech keyword-spotting algorithm. Without endpoint detection and division, we can get the minimum distance curve between continuous speech samples and every keyword-training net through the dynamic searching to the feature-extracted continuous speech. Then we can count the number of the keywords by investigating the vale-value and the numbers of the vales in the curve. Experiments of small vocabulary continuous speech with various speaking rate have got good recognition results and proved the validity of the algorithm.

Cognitive models in biomimetic pattern cognition

We describe a new model which is based on the concept of cognizing theory. The method identifies subsets of the data which are embedded in arbitrary oriented lower dimensional space. We definite k-mean covering, and study its property. Covering subsets of points are repeatedly sampled to construct trial geometry space of various dimensions. The sampling corresponding to the feature space having the best cognition ability between a mode near zero and the rest is selected and the data points are partitioned on the basis of the best cognition ability. The repeated sampling then continues recursively on each block of the data. We propose this algorithm based on cognition models. The experimental results for face recognition demonstrate that the correct rejection rate of the test samples excluded in the classes of training samples is very high and effective.

Study on text-dependent speaker recognition based on Biomimetic Pattern Recognition

We studied the application of Biomimetic Pattern Recognition to speaker recognition. A speaker recognition neural network using network matching degree as criterion is proposed. It has been used in the system of text-dependent speaker recognition. Experimental results show that good effect could be obtained even with lesser samples. Furthermore, the misrecognition caused by untrained speakers occurring in testing could be controlled effectively. In addition, the basic idea "cognition" of Biomimetic Pattern Recognition results in no requirement of retraining the old system for enrolling new speakers.

Dipole mode photonic crystal point defect laser on InGaAsP/InP

In this paper, we focus on the dipole mode of the two-dimensional (2D) photonic crystal (PC) single point defect cavity (SPDC) lasers and we report the fabrication and characterization of 2D PC SPDC lasers with the structure of adjusted innermost air holes. The photonic band and cavity Q factors are simulated by means of plane wave expansion (PWE) and finite-difference time-domain (FDTD), respectively. In order to improve the optical confinement of the SPDC, the diameter of the innermost holes was adjusted. Different lasing performances are observed experimentally. The experimental results agree with the theoretical prediction very well. (c) 2006 Elsevier B.V. All rights reserved.

A speaker-independent continuous speech recognition system using biomimetic pattern recognition

In speaker-independent speech recognition, the disadvantage of the most diffused technology (HMMs, or Hidden Markov models) is not only the need of many more training samples, but also long train time requirement. This paper describes the use of Biomimetic pattern recognition (BPR) in recognizing some mandarin continuous speech in a speaker-independent manner. A speech database was developed for the course of study. The vocabulary of the database consists of 15 Chinese dish's names, the length of each name is 4 Chinese words. Neural networks (NNs) based on Multi-weight neuron (MWN) model are used to train and recognize the speech sounds. The number of MWN was investigated to achieve the optimal performance of the NNs-based BPR. This system, which is based on BPR and can carry out real time recognition reaches a recognition rate of 98.14% for the first option and 99.81% for the first two options to the persons from different provinces of China speaking common Chinese speech. Experiments were also carried on to evaluate Continuous density hidden Markov models (CDHMM), Dynamic time warping (DTW) and BPR for speech recognition. The Experiment results show that BPR outperforms CDHMM and DTW especially in the cases of samples of a finite size.

Quantum measurement of single electron state by a quantum point contact

Closely related to the quantum information processing in solid states, we study the quantum measurement of single electron state by a mesoscopic charge-sensitive detector, namely the quantum point contact (QPC). We find that the conventional Lindblad-type master equation is not appropriate for describing the underlying measurement dynamics. The treatment developed in this work properly accounts for the energy-exchange between the detector and the measured system, and its role on the detailed-balance relation. A valid description for the QPC measurement dynamics is provided which may have impact on the study of quantum measurement and quantum feedback control in solid states.

Electronic structure and g factors of narrow-gap zinc-blende nanowires and nanorods

The Hamiltonian in the framework of eight-band effective-mass approximation of the zinc-blende nanowires and nanorods in the presence of external homogeneous magnetic field is given in the cylindrical coordinate. The electronic structure, optical properties, magnetic energy levels, and g factors of the nanowires and nanorods are calculated. It is found that the electron states consist of many hole-state components, due to the coupling of the conduction band and valence band. For the normal bands which are monotone functions of |k(z)|, long nanorods can be modeled by the nanowires, the energy levels of the nanorods approximately equal the values of the energy band E(k(z)) of the nanowires with the same radius at a special k(z), where k(z) is the wave vector in the wire direction. Due to the coupling of the states, some of the hole energy bands of the nanowires have their highest points at k(z)=0. Especially, the highest hole state of the InSb nanowires is not at the k(z)=0 point. It is an indirect band gap. For these abnormal bands, nanorods can not be modeled by the nanowires. The energy levels of the nanorods show an interesting plait-like pattern. The linear polarization factor is zero, when the aspect ratio L/2R is smaller than 1, and increases as the length increases. The g(z) and g(x) factors as functions of the k(z), radius R and length L are calculated for the wires and rods, respectively. For the wires, the g(z) of the electron ground state increases, and the g(z) of the hole ground state decreases first, then increases with the k(z) increasing. For the rods, the g(z) and g(x) of the electron ground state decrease as the R or the L increases. The g(x) of the hole ground state decreases, the g(z) of the hole ground state increases with the L increasing. The variation of the g(z) of the wires with the k(z) is in agreement with the variation of the g(z) of the rods with the L.

Research on multi-degree-of-freedom neurons with weighted graphs

In this paper, we redefine the sample points set in the feature space from the point of view of weighted graph and propose a new covering model - Multi-Degree-of-Freedorn Neurons (MDFN). Base on this model, we describe a geometric learning algorithm with 3-degree-of-freedom neurons. It identifies the sample points secs topological character in the feature space, which is different from the traditional "separation" method. Experiment results demonstrates the general superiority of this algorithm over the traditional PCA+NN algorithm in terms of efficiency and accuracy.

High-performance EML grown on taper-masked pattern substrates by ultra-low-pressure MOCVD

A novel in-plane bandgap energy controlling technique by ultra-low pressure (22 mbar) selective area growth (SAG) has been developed. To our knowledge, this is the lowest pressure condition during SAG process ever reported. In this work, high crystalline quality InGaAsP-InP MQWs with a photoluminescence (PL) full-width at half-maximum (FWHM) of less than 35meV are selectively grown on mask-patterned planar InP substrates by ultra-low pressure (22 mbar) metal-organic chemical vapor deposition (MOCVD). In order to study the uniformity of the MQWs grown in the selective area, novel tapered masks are designed and used. Through optimizing growth conditions, a wide wavelength shift of over 80 nm with a rather small mask width variation (0-30 mu m) is obtained. The mechanism of ultra-low pressure SAG is detailed by analyzing the effect of various mask designs and quantum well widths. This powerful technique is then applied to fabricate an electroabsorption-modulated laser (EML). Superior device characteristics are achieved, such as a low threshold current of 19mA and an output power of 7mW. (c) 2005 Elsevier B.V. All rights reserved.

Mode characteristics for equilateral triangle optical resonators

Mode characteristics of equilateral triangle resonators (ETRs) are analyzed based on the symmetry operation of the point group C-3v. The results show that doubly degenerate eigenstates can be reduced to the A(1) and A(2) representations of C-3v, if the longitudinal mode number is a multiple of 6; otherwise, they form the E irreducible representation Of C-3v. And the one-period length for the mode light ray is half of the perimeter of the ETR. Mode Q-factors are calculated by the finite-difference time-domain (FDTD) technique and compared with those calculated from far-field emission based on the analytical near-field pattern for TE and TM modes. The results show that the far-field emission based on the analytical field distribution can be used to estimate the mode Q-factor, especially for TM modes. FDTD numerical results also show that Q-factor of TE modes reaches maximum value as the longitudinal mode number is a multiple of 7. In addition, photoluminescence spectra and measured Q-factors are presented for fabricated ETR with side lengths of 20 and 30 mu m, and the mode wavelength intervals are compared with the analytical results.

A new development on ANN in China - Biomimetic pattern recognition and multi weight vector neurons

A new model of pattern recognition principles-Biomimetic Pattern Recognition, which is based on "matter cognition" instead of "matter classification", has been proposed. As a important means realizing Biomimetic Pattern Recognition, the mathematical model and analyzing method of ANN get breakthrough: a novel all-purpose mathematical model has been advanced, which can simulate all kinds of neuron architecture, including RBF and BP models. As the same time this model has been realized using hardware; the high-dimension space geometry method, a new means to analyzing ANN, has been researched.

Creation and suppression of point defects through a kick-out substitution process of Fe in InP

Indium antisite defect In-P-related photoluminescence has been observed in Fe-diffused semi-insulating (SI) InP. Compared to annealed undoped or Fe-predoped SI InP, there are fewer defects in SI InP obtained by long-duration, high-temperature Fe diffusion. The suppression of the formation of point defects in Fe-diffused SI InP can be explained in terms of the complete occupation by Fe at indium vacancy. The In-P defect is enhanced by the indium interstitial that is caused by the kick out of In and the substitution at the indium site of Fe in the diffusion process. Through these Fe-diffusion results, the nature of the defects in annealed undoped SI InP is better understood. (C) 2002 American Institute of Physics.

Point defects in III-V compound semiconductors

Point defects in III-V compound semiconductors were analyzed systematically in this paper. The effects of substitutes, antisites, interstitials, and vacancies on lattice parameters in III-V compound semiconductors were calculated with a simple model. The formation energies of vacancies in compound semiconductors can be obtained by this calculation. A practical technique established on this model has been utilized for measuring the stoichiometry in GaAs. The relationship between stoichiometry and deep level centers in GaAs was also investigated.