Geometrical learning, descriptive geometry, and biomimetic pattern recognition

Studies on learning problems from geometry perspective have attracted an ever increasing attention in machine learning, leaded by achievements on information geometry. This paper proposes a different geometrical learning from the perspective of high-dimensional descriptive geometry. Geometrical properties of high-dimensional structures underlying a set of samples are learned via successive projections from the higher dimension to the lower dimension until two-dimensional Euclidean plane, under guidance of the established properties and theorems in high-dimensional descriptive geometry. Specifically, we introduce a hyper sausage like geometry shape for learning samples and provides a geometrical learning algorithm for specifying the hyper sausage shapes, which is then applied to biomimetic pattern recognition. Experimental results are presented to show that the proposed approach outperforms three types of support vector machines with either a three degree polynomial kernel or a radial basis function kernel, especially in the cases of high-dimensional samples of a finite size. (c) 2005 Elsevier B.V. All rights reserved.

Correlation between Er3+ emission and Si clusters in Erbium-doped a-SiOx : H films

Erbium-doped hydrogenated amorphous silicon suboxide films containing silicon clusters (a-SiOx:H) were prepared. The samples exhibited photoluminescence (PL) peaks at around 750nm and 1.54 mu m, which could be assigned to the electron-hole recombination in silicon clusters and the intra-4f transition in Er3+, respectively. We compared annealing behaviors of Si clusters and Er3+ emission and found that Si clusters emission depends strongly upon crystallinity of Si clusters, whereas Er3+ emission is not sensitive to whether it is Si nanocrystals (nc-Si) or amorphous Si (a-Si) clusters. The erbium-doped a-SiOx:H films containing either a-Si clusters or nc-Si have the same kind of Er3+ -emitting centers. Based on these results, it is concluded that a-Si clusters can play the same role on Er3+ excitation as nc-Si. (c) 2004 Elsevier B.V. All rights reserved.

Manifold covering theory in biomimetic pattern recognition

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 manifold covering in biomimetic pattern recognition, and study its property. Furthermore, we propose this manifold covering algorithm based on Biomimetic Pattern Recognition. At last, the experimental results for face recognition demonstrates that the correct rejection rate of the test samples excluded in the classes of training samples is very high and effective.

Weak antilocalization and beating pattern in an InGaAs/InAlAs quantum well

We have observed the weak antilocalization (WAL) and beating SdH oscillation through magnetotransport measurements performed on a heavily delta-doped In0.52Al0.48As/In0.53Ga0.47As/In0.5Al0.48As single quantum well in an applied magnetic field up to 13 T and a temperature at 1.5 K. Both effects are caused by the strong Rashba spin-orbit (SO) coupling due to high structure inversion asymmetry (SIA). The Rashba SO coupling constant alpha and zerotield spin splitting Delta(0) are estimated and the obtained values are consistent from different analysis for this sample. (c) 2007 Elsevier Ltd. All rights reserved.

The correlation between preferred orientation and performance of ITO thin films

The tin-doped indium oxide (ITO) thin films were prepared by reactive thermal evaporation on the glass substrates. The effects of substrate temperatures (T-s) on the grain preferred orientation, the electrical and optical properties of ITO films were studied. X-ray diffraction (XRD) patterns indicated that the preferred orientation of film changes from (222) to (400) as T, > 200 degrees C. It can be explained by that the low-index crystallographic planes are easier to be formed when the adatoms have high surface mobility. The Hall measurements indicated that both the concentration and mobility of carrier increase with increasing T,,,. The grain orientation of film does not influence the transmissivity and the carrier concentration, but enhances the carrier mobility. The transmissivity of ITO films is over 90% in the visible wavelength region (except that of the film deposited at 125 degrees C). A minimum resistivity of 5 X 10-4 Omega cm is achieved for the (400) preferred orientation film. Thus, the highest figure of merit of 3.5 x 10(-2) square/Omega is obtained for the film with (400) preferred orientation. The correlation between the preferred orientation and electrical and optical properties are discussed.

Effect of electron-hole spatial correlation on spin relaxation dynamics in InAs submonolayer

Using time-resolved photoluminescence and time-resolved Kerr rotation spectroscopy, we explore the unique electron spin behavior in an InAs submonolayer sandwiched in a GaAs matrix, which shows very different spin characteristics under resonant and non-resonant excitations. While a very long spin relaxation lifetime of a few nanoseconds at low temperature is observed under non-resonant excitation, it decreases dramatically under resonant excitation. These interesting results are attributed to the difference in electron-hole interactions caused by non-geminate or geminate capture of photo-generated electron-hole pairs in the two excitation cases, and provide a direct verification of the electron-hole spatial correlation effect on electron spin relaxation. (c) 2007 Elsevier Ltd. All rights reserved.

Correlation between optical and structural properties of (Al,Ga)N layers grown by MOCVD

For both, (Al,Ga)N with low Al content grown on a GaN nucleation layer and (AI,Ga)N with high Al content gown on an AlN nucleation layer, the inhomogeneous distribution of the luminescence is linked to the distribution of defects, which may be due to inversion domains. In the former system, defect regions exhibit a much lower Al content than the nominal one leading to a splitting of the respective luminescence spectra. In the latter system, a domain-like growth is observed with a pyramidal surface morphology and non-radiative recombination within the domain boundaries. (c) 2007 WILEYNCH Verlag GmbH & Co. KGaA, Weinheim.

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.

Keyword spotting system based on biomimetic pattern recognition

The mandarin keyword spotting system was investigated, and a new approach was proposed based on the principle of homology continuity and point location analysis in high-dimensional space geometry theory which are both parts of biomimetic pattern recognition theory. This approach constructed a hyper-polyhedron with sample points in the training set and calculated the distance between each test point and the hyper-polyhedron. The classification resulted from the value of those distances. The approach was tested by a speech database which was created by ourselves. The performance was compared with the classic HMM approach and the results show that the new approach is much better than HMM approach when the training data is not sufficient.

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.

Correlation between optical and electrical properties in In0.52Al0.48As/InxGa1-xAs metamorphic high-electron-mobility-transistor structures on GaAs substrates

4.2 K photoluminescence (PL) and 77 K standard Hall-effect measurements were performed for In0.52Al0.48As/InxGa1-xAs metamorphic high-electron-mobility-transistor (HEMT) structures grown on GaAs substrates with different indium contents in the InxGa1-xAs well or different Si delta-doping concentrations. It was found that electron concentrations increased with increasing PL intensity ratio of the "forbidden" transition (the second electron subband to the first heavy-hole subband) to the sum of the "allowed" transition (the first electron subband to the first heavy-hole subband) and the forbidden transition. And electron mobilities decreased with increasing product of the average full width at half maximum of allowed and forbidden transitions and the electron effective mass in the InxGa1-xAs quantum well. These results show that PL measurements are a good supplemental tool to Hall-effect measurements in optimization of the HEMT layer structure. (c) 2006 American Institute of Physics.

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.

Raman evidence for atomic correlation between the two constituent tubes in double-walled carbon nanotubes

Four well-resolved peaks with very narrow linewidths were found in the D-band and G'-band features of double-walled carbon nanotubes (DWNTs). This fact implies the occurrence of additional van Hove singularities (vHSs) in the joint density of states (JDOS) of DWNTs, which is consistent with theoretical calculations. According to their peak frequencies and theoretical analysis, the two outer peaks can be deduced to originate from a strong coupling between the two constituent tubes of commensurate DWNTs and the two inner peaks were curvature-related and assigned to originate from the two tubes with a weak coupling. This observation and elucidation constitute the first Raman evidence for atomic correlation and the resulting electronic structure change of the two constituent tubes in DWNTs. This result opens the possibility of predicting and modifying the electronic properties of DWNTs for their electronic applications.