A method of Biomimetic Pattern Recognition for face recognition

A new method of face recognition, based on Biomimetic Pattern Recognition and Multi-Weights Neuron Network, had been proposed. A model for face recognition that is based on Biomimetic Pattern Recognition had been discussed, and a new method of facial feature extraction also had been introduced. The results of experiments with BPR and K-Nearest Neighbor Rules showed that the method based on BPR can eliminate the error recognition of the samples of the types that not be trained, the correct rate is also enhanced.

Biomimetic (topological) pattern recognition - A new model of pattern recognition theory and its application

A new theoretical model of Pattern Recognition principles was proposed, which is based on "matter cognition" instead of "matter classification" in traditional statistical Pattern Recognition. This new model is closer to the function of human being, rather than traditional statistical Pattern Recognition using "optimal separating" as its main principle. So the new model of Pattern Recognition is called the Biomimetic Pattern Recognition (BPR)(1). Its mathematical basis is placed on topological analysis of the sample set in the high dimensional feature space. Therefore, it is also called the Topological Pattern Recognition (TPR). The fundamental idea of this model is based on the fact of the continuity in the feature space of any one of the certain kinds of samples. We experimented with the Biomimetic Pattern Recognition (BPR) by using artificial neural networks, which act through covering the high dimensional geometrical distribution of the sample set in the feature space. Onmidirectionally cognitive tests were done on various kinds of animal and vehicle models of rather similar shapes. For the total 8800 tests, the correct recognition rate is 99.87%. The rejection rate is 0.13% and on the condition of zero error rates, the correct rate of BPR was much better than that of RBF-SVM.

Silicon: the promising material for phontonic integrated circuits platform

The development of optical network demands integrated arid multiple functionality modules to lowing cost and acquire highly reliability. Among the various contender materials to be photonic integrated circuits platform, silicon exhibits dominant characteristics and is the most promising platform materials. The paper compares the characteristics of some candidate materials with silicon and reviews recent progress in silicon based photonic integration technology. Tile challenges to silicon for optical integration for optical networking application arc also indicated.

Growth and fabrication of high performance 980nm strained InGaAs quantum well lasers using novel hybrid material system of InGaAsP and AlGaAs

In this paper, we report on the design, growth and fabrication of 980nm strained InGaAs quantum well lasers employing novel material system of Al-free active region and AlGaAs cladding layers. The use of AlGaAs cladding instead of InGaP provides potential advantages in laser structure design, improvement of surface morphology and laser performance. We demonstrate an optimized broad-waveguide structure for obtaining high power 980nm quantum well lasers with low vertical beam divergence. The laser structure was grown by low-pressure metalorganic chemical vapor deposition, which exhibit a high internal quantum efficiency of similar to 90% and a low internal loss of 1.5-2.5 cm(-1). The broad-area and ridge-waveguide laser devices are both fabricated. For 100 mu m wide stripe lasers with cavity length of 800 mu m, a low threshold current of 170mA, a high slope efficiency of 1.0W/A and high output power of more than 3.5W are achieved. The temperature dependences of the threshold current and the emitting spectra demonstrate a very high characteristic temperature coefficient (T-o) of 200-250K and a wavelength shift coefficient of 0.34nm/degrees C. For 4 mu m-width ridge waveguide structure laser devices, a maximum output power of 340mW with GOD-free thermal roll-over characteristics is obtained.

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.

Heteroepitaxial growth of novel SOI material Si/gamma-Al2O3/Si

In this paper, we report the fabrication of Si-based double hetero-epitaxial SOI materials Si/gamma-Al2O3/Si. First, single crystalline gamma-Al2O3 (100) insulator films were grown epitaxially on Si(100) by LPCVD, and then, Si(100) epitaxial films were grown on gamma-Al2O3 (100)/Si(100) epi-substrates using a CVD method similar to silicon on sapphire (SOS) epitaxial growth. The Si/gamma-Al2O3 (100)/Si(100) SOI materials are characterized in detail by RHEED, XRD and AES techniques. The results demonstrate that the device-quality novel SOI materials Si/gamma-Al2O3 (100)/Si(100) has been fabricated successfully and can be used for application of MOS device.

Improved neutron radiation hardness for Si detectors: Application of low resistivity starting material and or manipulation of N-eff by selective filling of radiation-induced traps at low temperatures

Radiation-induced electrical changes in both space charge region (SCR) of Si detectors and bulk material (BM) have been studied for samples of diodes and resistors made on Si materials with different initial resistivities. The space charge sign inversion fluence (Phi(inv)) has been found to increase linearly with the initial doping concentration (the reciprocal of the resistivity), which gives improved radiation hardness to Si detectors fabricated from low resistivity material. The resistivity of the BM, on the other hand, has been observed to increase with the neutron fluence and approach a saturation value in the order of hundreds k Omega cm at high fluences, independent of the initial resistivity and material type. However, the fluence (Phi(s)), at which the resistivity saturation starts, increases with the initial doping concentrations and the value of Phi(s) is in the same order of that of Phi(inv) for all resistivity samples. Improved radiation hardness can also be achieved by the manipulation of the space charge concentration (N-eff) in SCR, by selective filling and/or freezing at cryogenic temperatures the charge state of radiation-induced traps, to values that will give a much smaller full depletion voltage. Models have been proposed to explain the experimental data.

Material transport in self-assembled InAs/GaAs quantum dot ensemble

Introducing the growth interruption between the InAs deposition and subsequent GaAs growth in self-assembled quantum dot (QD) structures, the material transport process in the InAs layers has been investigated by photoluminescence and transmission electron microscopy measurement. InAs material in structures without misfit dislocations transfers from the wetting layer to QDs corresponding to the red-shift of PL peak energy due to interruption. On the other hand, the PL peak shifts to higher energy in the structures with dislocations. In this case, the misfit dislocations would capture the InAs material from the surrounding wetting layer and coherent islands leading to the reduction of the size of these QDs. The variations in the PL intensity and Linewidth are also discussed.