Computer generated hologram from point cloud using graphics processor

Computer generated holography is an extremely demanding and complex task when it comes to providing realistic reconstructions with full parallax, occlusion, and shadowing. We present an algorithm designed for data-parallel computing on modern graphics processing units to alleviate the computational burden. We apply Gaussian interpolation to create a continuous surface representation from discrete input object points. The algorithm maintains a potential occluder list for each individual hologram plane sample to keep the number of visibility tests to a minimum.We experimented with two approximations that simplify and accelerate occlusion computation. It is observed that letting several neighboring hologramplane samples share visibility information on object points leads to significantly faster computation without causing noticeable artifacts in the reconstructed images. Computing a reduced sample set via nonuniform sampling is also found to be an effective acceleration technique. © 2009 Optical Society of America.

TUNPOP, a computer simulation model of the yellowfin tuna population and the surface tuna fishery of the Eastern Pacific Ocean

ENGLISH: Mathematical documentation of TUNP0P, an age-structured computer simulation model of the yellowfin tuna population and surface tuna fishery of the eastern Pacific Ocean, is described. Example runs of the model are presented and discussed, and the sensitivity of the model output to changes in various parameters is examined. SPANISH: Se describe la documentación matemática de TUNP0P, un modelo computador de simulación basado en la edad de la población del atún aleta amarilla y de la pesca atunera epipelágíca del Océano Pacífico oriental. Se presentan y se discuten ejemplos de las pasadas del modelo, y se examina la sensibilidad de los resultados de salida con relación a los cambios de varios parámetros. (PDF contains 47 pages.)

Síntese evolucionária de circuitos sequenciais inspirada nos princípios da computação quântica.

Esta dissertação investiga a aplicação dos algoritmos evolucionários inspirados na computação quântica na síntese de circuitos sequenciais. Os sistemas digitais sequenciais representam uma classe de circuitos que é capaz de executar operações em uma determinada sequência. Nos circuitos sequenciais, os valores dos sinais de saída dependem não só dos valores dos sinais de entrada como também do estado atual do sistema. Os requisitos cada vez mais exigentes quanto à funcionalidade e ao desempenho dos sistemas digitais exigem projetos cada vez mais eficientes. O projeto destes circuitos, quando executado de forma manual, se tornou demorado e, com isso, a importância das ferramentas para a síntese automática de circuitos cresceu rapidamente. Estas ferramentas conhecidas como ECAD (Electronic Computer-Aided Design) são programas de computador normalmente baseados em heurísticas. Recentemente, os algoritmos evolucionários também começaram a ser utilizados como base para as ferramentas ECAD. Estas aplicações são referenciadas na literatura como eletrônica evolucionária. Os algoritmos mais comumente utilizados na eletrônica evolucionária são os algoritmos genéticos e a programação genética. Este trabalho apresenta um estudo da aplicação dos algoritmos evolucionários inspirados na computação quântica como uma ferramenta para a síntese automática de circuitos sequenciais. Esta classe de algoritmos utiliza os princípios da computação quântica para melhorar o desempenho dos algoritmos evolucionários. Tradicionalmente, o projeto dos circuitos sequenciais é dividido em cinco etapas principais: (i) Especificação da máquina de estados; (ii) Redução de estados; (iii) Atribuição de estados; (iv) Síntese da lógica de controle e (v) Implementação da máquina de estados. O Algoritmo Evolucionário Inspirado na Computação Quântica (AEICQ) proposto neste trabalho é utilizado na etapa de atribuição de estados. A escolha de uma atribuição de estados ótima é tratada na literatura como um problema ainda sem solução. A atribuição de estados escolhida para uma determinada máquina de estados tem um impacto direto na complexidade da sua lógica de controle. Os resultados mostram que as atribuições de estados obtidas pelo AEICQ de fato conduzem à implementação de circuitos de menor complexidade quando comparados com os circuitos gerados a partir de atribuições obtidas por outros métodos. O AEICQ e utilizado também na etapa de síntese da lógica de controle das máquinas de estados. Os circuitos evoluídos pelo AEICQ são otimizados segundo a área ocupada e o atraso de propagação. Estes circuitos são compatíveis com os circuitos obtidos por outros métodos e em alguns casos até mesmo superior em termos de área e de desempenho, sugerindo que existe um potencial de aplicação desta classe de algoritmos no projeto de circuitos eletrônicos.

Recurrent input transformations for Hidden Markov models

This paper presents a new architecture which integrates recurrent input transformations (RIT) and continuous density HMMs. The basic HMM structure is extended to accommodate recurrent neural networks which transform the input observations before they enter the Gaussian output distributions associated with the states of the HMM. During training the parameters of both HMM and RIT are simultaneously optimized according to the Maximum Mutual Information (MMI) criterion. Results are presented for the E-set recognition task which demonstrate the ability of recurrent input transformations to exploit longer term correlations in the speech signal and to give improved discrimination.

Compact holographic optical interconnections using nematic liquid crystal spatial light modulators

In a fibre-optic communication network, the wavelength-division multiplexing (WDM) technique enables an expansion of the data-carrying capacity of optical fibres. This can be achieved by transmitting different channels on a single optical fibre, with each channel modulating a different wavelength. In order to access and manipulate these channels at a node of the network, a compact holographic optical switch is designed, modelled, and constructed. The structure of such a switch consists of a series of optical components which are used to collimate the beam from the input, de-multiplex each individual wavelength into separated channels, manipulate the separated channels, and reshape the beam to the output. A spatial light modulator (SLM) is crucial in this system, offering control and flexibility at the channel manipulation stage, and providing the ability to redirect light into the desired output fibre. This is achieved by the use of a 2-D analogue phase computer generated hologram (CGH) based on liquid crystal on silicon (LCOS) technology. © 2011 Copyright Society of Photo-Optical Instrumentation Engineers (SPIE).

SoC energy savings = reduce+reuse+recycle: A case study using a 660MHz DC-DC converter with integrated output filter

This paper advocates 'reduce, reuse, recycle' as a complete energy savings strategy. While reduction has been common to date, there is growing need to emphasize reuse and recycling as well. We design a DC-DC buck converter to demonstrate the 3 techniques: reduce with low-swing and zero voltage switching (ZVS), reuse with supply stacking, and recycle with regulated delivery of excess energy to the output load. The efficiency gained from these 3 techniques helps offset the loss of operating drivers at very high switching frequencies which are needed to move the output filter completely on-chip. A prototype was fabricated in 0.18μm CMOS, operates at 660MHz, and converts 2.2V to 0.75-1.0V at ∼50mA.1 © 2008 IEEE.

Gaussian Process training with input noise

In standard Gaussian Process regression input locations are assumed to be noise free. We present a simple yet effective GP model for training on input points corrupted by i.i.d. Gaussian noise. To make computations tractable we use a local linear expansion about each input point. This allows the input noise to be recast as output noise proportional to the squared gradient of the GP posterior mean. The input noise variances are inferred from the data as extra hyperparameters. They are trained alongside other hyperparameters by the usual method of maximisation of the marginal likelihood. Training uses an iterative scheme, which alternates between optimising the hyperparameters and calculating the posterior gradient. Analytic predictive moments can then be found for Gaussian distributed test points. We compare our model to others over a range of different regression problems and show that it improves over current methods.

Toward automated generation of parametric BIMs based on hybrid video and laser scanning data

Only very few constructed facilities today have a complete record of as-built information. Despite the growing use of Building Information Modelling and the improvement in as-built records, several more years will be required before guidelines that require as-built data modelling will be implemented for the majority of constructed facilities, and this will still not address the stock of existing buildings. A technical solution for scanning buildings and compiling Building Information Models is needed. However, this is a multidisciplinary problem, requiring expertise in scanning, computer vision and videogrammetry, machine learning, and parametric object modelling. This paper outlines the technical approach proposed by a consortium of researchers that has gathered to tackle the ambitious goal of automating as-built modelling as far as possible. The top level framework of the proposed solution is presented, and each process, input and output is explained, along with the steps needed to validate them. Preliminary experiments on the earlier stages (i.e. processes) of the framework proposed are conducted and results are shown; the work toward implementation of the remainder is ongoing.

Expressive visual text-to-speech using active appearance models

This paper presents a complete system for expressive visual text-to-speech (VTTS), which is capable of producing expressive output, in the form of a 'talking head', given an input text and a set of continuous expression weights. The face is modeled using an active appearance model (AAM), and several extensions are proposed which make it more applicable to the task of VTTS. The model allows for normalization with respect to both pose and blink state which significantly reduces artifacts in the resulting synthesized sequences. We demonstrate quantitative improvements in terms of reconstruction error over a million frames, as well as in large-scale user studies, comparing the output of different systems. © 2013 IEEE.

Method for improving classification performance of neural network based on fuzzy input and network inversion

In order to effectively improve the classification performance of neural network, first architecture of fuzzy neural network with fuzzy input was proposed. Next a cost function of fuzzy outputs and non-fuzzy targets was defined. Then a learning algorithm from the cost function for adjusting weights was derived. And then the fuzzy neural network was inversed and fuzzified inversion algorithm was proposed. Finally, computer simulations on real-world pattern classification problems examine the effectives of the proposed approach. The experiment results show that the proposed approach has the merits of high learning efficiency, high classification accuracy and high generalization capability.

Influence of lateral propagating modes on laser output characteristics in selectively oxidized vertical cavity surface-emitting lasers with double oxide layers

The influence of lateral propagating modes on the threshold current and the spontaneous emission factor in selectively oxidized vertical cavity surface-emitting lasers (VCSELs) is investigated based on the mode behaviors of lateral propagating modes and the rate equation model. The numerical results show that the lateral propagating modes may be trapped in the aperture region for the selectively oxidized VCSEL with two oxide layers, one above and one below the active region. The output characteristics of VCSELs can be affected due to the reabsorption of the quasitrapped lateral propagating modes. A lower threshold current can be expected for a VCSEL with double oxide layers than that with a single oxide layer. The numerical results of rate equations also show that a larger spontaneous emission factor can be obtained by fitting the output-input curves for the VCSEL with double oxide layers. (C) 1999 American Institute of Physics. [S0021-8979(99)07919-0].

A novel low-power input-independent MOS AC/DC charge pump

This paper presents a novel fully integrated MOS AC to DC charge pump with low power dissipation and stable output for RFID applications. To improve the input sensitivity, we replaced Schottky-diodes in conventional charge pumps with MOS diodes with zero threshold, which has less process defects and is thus more compatible with other circuits. The charge pump in a RFID transponder is implemented in a 0.35um CMOS technology with 0.24 sq mm die size. The analytical model of the charge pump and the simulation results are presented.

A Simplified Soft-output QRD-M MIMO Detector

In this paper, a low-complexity soft-output QRD-M detection algorithm is proposed for high-throughput Multiple-input multiple-output (MIMO) systems. By employing novel expansion on demand and distributed sorting scheme, the proposed algorithm can reduce 70% and 85% foundational operations for 16-QAM and 64-QAM respectively compared to the conventional QRD-M algorithm. Furthermore, the proposed algorithm can yield soft information to improve the bit error rate (BER) performance. Simulation results show that the proposed algorithm can achieve a near-NIL detection performance with less foundational operations

A COMMON-CAVITY 2-SECTION INGAASP/INP BISTABLE LASER WITH A LOW OPTICAL SWITCHING POWER

The characteristics of the steady-state and the transient response to external light excitation of a common-cavity two-section (CCTS) bistable semiconductor laser is investigated. The results on the relation of light output versus light input, the wavelength match, optical amplification and optical switching are presented. Experimental results are compared to the results of a computer simulation.