Time filter

Source Type

Silver Springs, NV, United States

Noguchi K.,Kanazawa Institute of Technology | Rajagopalan H.,University of California at Los Angeles | Rajagopalan H.,Apple USA | Rahmat-Samii Y.,Kanazawa Institute of Technology
IEEE Transactions on Antennas and Propagation | Year: 2016

This paper proposes a new circuit model for E-shaped patch antennas (ESPA) using the multiconductor transmission line mode theory (the modal theory). First, radiation and transmission line modes generated on the ESPA are described, and an equivalent circuit is derived from the modal theory. The equivalent circuit is analyzed in detail to obtain wideband and multiband characteristics. For wideband ESPAs, the theoretical maximum bandwidth is derived under VSWR criterion. Dual-band ESPA is also discussed theoretically. Finally, impedance characteristics obtained by the circuit model are compared with full-wave electromagnetic simulations and measurements. © 2015 IEEE.

Swietojanski P.,University of Edinburgh | Ghoshal A.,University of Edinburgh | Ghoshal A.,Apple USA | Renals S.,University of Edinburgh
IEEE Signal Processing Letters | Year: 2014

We investigate convolutional neural networks (CNNs) for large vocabulary distant speech recognition, trained using speech recorded from a single distant microphone (SDM) and multiple distant microphones (MDM). In the MDM case we explore a beamformed signal input representation compared with the direct use of multiple acoustic channels as a parallel input to the CNN. We have explored different weight sharing approaches, and propose a channel-wise convolution with two-way pooling. Our experiments, using the AMI meeting corpus, found that CNNs improve the word error rate (WER) by 6.5% relative compared to conventional deep neural network (DNN) models and 15.7% over a discriminatively trained Gaussian mixture model (GMM) baseline. For cross-channel CNN training, the WER improves by 3.5% relative over the comparable DNN structure. Compared with the best beamformed GMM system, cross-channel convolution reduces the WER by 9.7% relative, and matches the accuracy of a beamformed DNN. © 1994-2012 IEEE.

Bharwani A.M.,University of Calgary | Harris G.C.,Apple USA | Southwick F.S.,Florida College
Academic Medicine | Year: 2012

An effective interprofessional medical team can efficiently coordinate health care providers to achieve the collective outcome of improving each patient's health. To determine how current teams function, four groups of business students independently observed interprofessional work rounds on four different internal medicine services in a typical academic hospital and also interviewed the participants. In all instances, caregivers had formed working groups rather than working teams. Participants consistently exhibited parallel interdependence (individuals working alone and assuming their work would be coordinated with other caregivers) rather than reciprocal interdependence (individuals working together to actively coordinate patient care), the hallmark of effective teams. With one exception, the organization was hierarchical, with the senior attending physician possessing the authority. The interns exclusively communicated with the attending physician in one-on-one conversations that excluded all other members of the team. Although nurses and pharmacists were often present, they never contributed their ideas and rarely spoke.The authors draw on these observations to form recommendations for enhancing interprofessional rounding teams. These are to include the bedside nurse, pharmacist, and case manager as team members, begin with a formal team launch that encourages active participation by all team members, use succinct communication protocols, conduct work rounds in a quiet, distraction-free environment, have teams remain together for longer durations, and receive teamwork training and periodic coaching. High-performing businesses have effectively used teams for decades to achieve their goals, and health care professionals should follow this example.

Raguram R.,Apple USA | Chum O.,Czech Technical University | Pollefeys M.,ETH Zurich | Matas J.,Czech Technical University | Frahm J.-M.,University of North Carolina at Chapel Hill
IEEE Transactions on Pattern Analysis and Machine Intelligence | Year: 2013

A computational problem that arises frequently in computer vision is that of estimating the parameters of a model from data that have been contaminated by noise and outliers. More generally, any practical system that seeks to estimate quantities from noisy data measurements must have at its core some means of dealing with data contamination. The random sample consensus (RANSAC) algorithm is one of the most popular tools for robust estimation. Recent years have seen an explosion of activity in this area, leading to the development of a number of techniques that improve upon the efficiency and robustness of the basic RANSAC algorithm. In this paper, we present a comprehensive overview of recent research in RANSAC-based robust estimation by analyzing and comparing various approaches that have been explored over the years. We provide a common context for this analysis by introducing a new framework for robust estimation, which we call Universal RANSAC (USAC). USAC extends the simple hypothesize-and-verify structure of standard RANSAC to incorporate a number of important practical and computational considerations. In addition, we provide a general-purpose C++ software library that implements the USAC framework by leveraging state-of-the-art algorithms for the various modules. This implementation thus addresses many of the limitations of standard RANSAC within a single unified package. We benchmark the performance of the algorithm on a large collection of estimation problems. The implementation we provide can be used by researchers either as a stand-alone tool for robust estimation or as a benchmark for evaluating new techniques. © 1979-2012 IEEE.

Kovitz J.M.,University of California at Los Angeles | Rajagopalan H.,University of California at Los Angeles | Rajagopalan H.,Apple USA | Rahmat-Samii Y.,University of California at Los Angeles
IEEE Antennas and Wireless Propagation Letters | Year: 2012

Bias lines form an integral part of the reconfigurable antenna design process. Current methodologies for bias lines are either too narrowband or require a relatively large amount of real-estate that may not be available. Also, metallic lines are unsuitable for designs with strong fields in the bias line location. Bias lines using resistive materials are often prototyped using microfabrication facilities, but these facilities might not always be accessible. This letter investigates a novel, practical, and cost-effective bias line solution using conductive adhesives. These resistive lines effectively attenuate the RF signals providing good isolation between the RF and the dc signals while still passing a dc voltage to activate the switches. The reconfigurable E-shaped patch antenna is used as a case study in this letter, and some issues when using metallic lines in this type of antenna are also presented. The newly proposed bias line aims to minimize those issues shown hereafter. The fabrication process is also enumerated for those interested in repeating these designs for other applications. Overall, the S11 and pattern measurements show good agreement with the simulations and prove their effectiveness experimentally. © 2002-2011 IEEE.

Discover hidden collaborations