On body-fitted co-ordinates for separated laminar flows past wall-mounted obstacles

A body-fitted curvilinear co-ordinate system is used to solve the equations of two-dimensional incompressible laminar flow over bluff obstructions by finite differences. Arbitrary conditions at the corner are removed by this method. Results for a backward-facing step are in reasonable agreement with those obtained with conventional mesh systems, and the differences are explained. A treatment of a channel expansion, in comparison with empirical data, is also included. The capability of the present method to handle arbitrary two-dimensional geometries is stressed and demonstrated, using a triangle and a semi-circle as examples.     

Status of Work on the VÉPP-5 Injection Complex

The VÉPP-5 injection complex under construction at the Institute of Nuclear Physics of the Siberian Branch of the Russian Academy of Sciences is a powerful source of intense electron and positron bunches at 510 MeV, which covers all needs of the electron–positron colliding beam setups currently operating and under construction at the Institute of Nuclear Physics. The complex includes a 285 MeV linear electron accelerator, a 510 MeV linear positron accelerator, and an accumulator–cooler with beam injection and ejection channels. Intense work on the design, assembly, and tuning of the linear electron accelerator has been conducted in the last 2 yr. As a result, by August 2002 the linear electron accelerator was put into operation with all standard subsystems. By this time, the isochronous achromatic turning of the electron beam, a system for converting electrons into positrons, and the first accelerating structure of the linear positron accelerator were assembled and put into operation. All this made it possible to accelerate the positron beam up to 75 MeV. Preliminary results of tests of the linear accelerators are presented.     

Computation of 2D discrete Hartley transform

Based on a decimation-in-time decomposition, a fast split-radix algorithm for the 2D discrete Hartley transform is presented. Compared to other reported algorithms, the proposed algorithm achieves substantial savings on the number of operations and provides a wider choice of transform sizes     

Eigenfeature regularization and extraction in face recognition

This work proposes a subspace approach that regularizes and extracts eigenfeatures from the face image. Eigenspace of the within-class scatter matrix is decomposed into three subspaces: a reliable subspace spanned mainly by the facial variation, an unstable subspace due to noise and finite number of training samples and a null subspace. Eigenfeatures are regularized differently in these three subspaces based on an eigenspectrum model to alleviate problems of instability, over-fitting or poor generalization. This also enables the discriminant evaluation performed in the whole space. Feature extraction or dimensionality reduction occurs only at the final stage after the discriminant assessment. These efforts facilitate a discriminative and stable low-dimensional feature representation of the face image. Experiments comparing the proposed approach with some other popular subspace methods on the FERET, ORL, AR and GT databases show that our method consistently outperforms others.     

Complete discriminant evaluation and feature extraction in kernel space for face recognition

This work proposes a method to decompose the kernel within-class eigenspace into two subspaces: a reliable subspace spanned mainly by the facial variation and an unreliable subspace due to limited number of training samples. A weighting function is proposed to circumvent undue scaling of eigenvectors corresponding to the unreliable small and zero eigenvalues. Eigenfeatures are then extracted by the discriminant evaluation in the whole kernel space. These efforts facilitate a discriminative and stable low-dimensional feature representation of the face image. Experimental results on FERET, ORL and GT databases show that our approach consistently outperforms other kernel based face recognition methods.

Improved photon management in a photoelectrochemical cell with Nd-modified TiO2 thin film photoanode

Photon management involving particularly an up-conversion process is proposed as a relatively novel strategy for improving the efficiency of hydrogen generation in photoelectrochemical cells (PEC) with wide-band gap photoanodes. Optically active photoanode has been constructed by electrodeposition of titanium dioxide nanopowders containing Nd³⁺ ions, synthesized via a sol-gel method, onto ITO/TiO₂(thin film) substrates. Thin films of TiO₂ have been deposited by means of RF magnetron sputtering in an ultra-high-vacuum system. X-ray diffraction, scanning electron microscopy, UV-VIS-NIR spectrophotometry, and photoluminescence have been applied to assess the properties of photoanodes. In experiments involving photon-assisted water splitting, an external up-converter containing Yb³⁺/Er³⁺ rare-earth ions has been used. Photocurrent as a function of voltage (V_B) under illumination with white light is relatively high (280 μA at V_B = 0 V) for pure TiO₂ thin films and it is not affected by the electrodeposition of TiO₂:Nd³⁺ powders. NIR-driven up-conversion with laser excitation at λ = 980 nm has been found responsible for a 13-fold increase in photocurrent at V_B = 0 V in the modified PEC configuration.