No description available.

京都大学 / 新制・論文博士 / 博士(人間健康科学) / 乙第13483号 / 論人健博第9号 / 新制||人健||7(附属図書館) / 京都大学大学院医学研究科人間健康科学系専攻 / (主査)教授 藤井 康友, 教授 澤本 伸克, 教授 長尾 美紀 / 学位規則第4条第2項該当 / Doctor of Human Health Sciences / Kyoto University / DFAM

Abstract A Keyword Spotting System (KWS) is a system that recognizes predefined keywords in spoken utterances or written documents. The objective is to obtain the highest possible keyword detection rate without increasing the number of false detections in a system. The common approach to keyword spotting is the use of a Hidden Markov Model (HMM). These are usually complex systems which require training speech data. The Typical HMM approach uses garbage templates or HMM models to match non-keyword speech and non-speech sounds. The purpose of this research is to design a simple Keyword Spotting System. The system will be designed to spot English words and should be easily adaptable to other languages There are many challenges in designing a keyword spotting system such as variations in speech like pitch, loudness, timbre that make recognition difficult. There can be wide variations in utterances even from the same speaker. In this research, the use of cross-correlation, as an alternative means for detecting keywords in an utterance, was investigated. This research also involves the modeling of a global keyword using a quantized dynamic time warping algorithm, which can function effectively with multi-speakers. The global keyword is an aggregation of the features from several occurrences of the same keyword. This research also investigates the effect of pitch normalization on keyword detection. The use of cross-correlation as a method for keyword spotting was investigated in both the time and MFCC domain. In the time domain the global keyword was cross-correlated with a pitch-normalized utterance. A zero lag ratio (the ratio of the power around the zero lag obtained from a cross correlation to the power in the rest of the signal is computed) was computed for each speech frame, a threshold was then used to determine if the keyword is present. For the MFCC domain the MFCC features of each keyword were computed, normalized and cross-correlated with the normalized MFCC features of portions of the utterance of the same size as the keyword. Cross-correlation of MFCC features of the keyword with that of each portion of the utterance yields a single value between 0-1. The portion with the highest value is usually the location of the keyword. Results in the time domain varied from keyword to keyword, some words showed a 60% hit rate while the average obtained from various keywords from the Call Home database had an average of 41%. Cross-correlation of the keywords and utterance in the MFCC domain yielded a 66% hit rate in test conducted on all different keywords in the Call Home and Switchboard corpus. The system accuracy is keyword dependent with some keywords having an 85% hit rate / Electrical and Computer Engineering

Particle image velocimetry (PIV) has become one of the most important tools for experimentally investigating the physics of fluid flows. In PIV, image-processing algorithms estimate flow velocity by measuring the displacements of flow-tracer particles suspended in a fluid. The fundamental operation in PIV is the cross correlation (CC), which measures the displacement between two similar patterns. These measurements can fail under circumstances that arise due to the nature of the underlying flow field (e.g., vortices and boundary layers, where particle patterns not only translate but also rotate, stretch, and shear) or of the images (e.g., X-ray images, with comparatively low signal to noise ratios). Despite these shortcomings, fairly little attention has been paid to fundamentally improving measurements at the level of the CC. The objective of this dissertation is to demonstrate specific modifications to the correlation kernel of PIV that increase its accuracy and in certain cases extend its utility to classes of flows and image types that were previously unresolvable. First, we present a new PIV correlation algorithm called the Fourier-Mellin correlation (FMC) that reduces velocity errors by an order of magnitude in rotating flows (chapter 1). Second, we develop a model of PIV cross correlations that explains the fundamental sources of several major drivers of error in these measurements. We show how the shapes of the tracer particles and the distributions of their individual displacements affect the correlation signal to noise ratio (SNR), whose effects have previously been described only heuristically. We use this insight to create an algorithm that automatically creates a Fourier-based weighting filter, and demonstrate that our algorithm reduces bias and RMS errors in multiple types of PIV experiments (chapter 2). Finally, we apply principles from our insights to measure blood flows in the hearts of grasshoppers using X-ray PIV, and discovered flow kinematics that were unexpected according to the current prevailing understanding of the heart as a peristaltic pump that produces directional flows. Our results suggest that flow production in insect hearts may be more complex than once thought (chapter 3). / Ph. D. / Particle image velocimetry (PIV) is a tool for measuring the motion of fluid flows. In PIV, reflective particles are suspended in a flowing fluid, and cameras record their motion. Computer algorithms measure the motion of the particles in those images to estimate the velocity of the fluid. This dissertation is about the theory, algorithms, and experiments of particle image velocimetry. We explain from a theoretical standpoint the reasons that PIV can fail to provide reliable measurements for several types of flows that are commonly encountered in the research of fluid physics and engineering, such as swirling vortices or eddies, jets, turbulence, and microscopic flows. We apply this understanding to create new algorithms that improve PIV measurements in these kinds of challenging scenarios. Lastly, we use PIV and high-speed Xray imaging to measure flow patterns within the tubular hearts of living grasshoppers. From these experiments, we discovered flow behaviors that were strikingly different from what we expected according to the current prevailing presumption that the insect heart is a peristaltic pump. If the heart is proven to function other than by peristalsis, then this could imply that a previously overlooked flow mechanism could in fact be among the most prevalent among animals.

A study was made in order to deduce whether certain chemical species, namely water (H2O) and carbon monoxide (CO), are present in the atmosphere of the exoplanet WASP-107 b, which lies about 200 light-years away from Earth in the constellation Virgo. The project was carried out at Uppsala University at the Department of Physics and Astronomy. This was done through the use of transmission spectroscopy and executed using a cross-correlation technique, one of the leading methods available today to extract exoplanetary atmospheric information. The data used was collected during a transit which occured in March 2022, originally gathered by the spectrograph CRIRES+ stationed at Very Large Telescope (VLT) in Paranal. WASP-107 b is a warm Jupiter-type planet, and since the aforementioned chemical species exhibit spectral lines mainly in the infrared (0.95-5.3 μm), it makes CRIRES+ a desirable instrument due to its specialization for working in the infrared. The data analysis was performed using several scripts built in Python with subsequent data-reduction methods. The data-reduction methods used for this purpose was the standard ESO CRIRES+ data reduction pipeline which includes removal of systemic sources of noise such as dead pixels and cosmic rays, and SysRem, which is an algorithm used to remove any trends with time and any constant features in time for each pixel time series. SysRem is currently one of the most efficient way available for doing so, and is commonly used in these types of studies. Several detection maps were then generated and studied in order to deduce whether a detection had been made or not. For this project, one exoplanet was examined and its atmosphere was probed for H2O and CO. The cross-correlation templates utilized were a combination of both species as well as one corresponding to only CO. The detection maps generated from the cross-correlation analysis initially suggested non-detections for all combinations of SysRem iterations and templates, except for two which presented features that might imply a detection but without any strong certainty. Those results indicate the possible existence of CO in the atmosphere of WASP-107 b, but further investigation is needed in order to determine their validity. / Denna rapport beskriver en studie som utförts för att undersöka ifall vissa kemiska arter, nämligen vatten (H2O) och kolmonoxid (CO), existerar i atmosfären kring exoplaneten WASP-107 b. Exoplaneten ligger cirka 200 ljusår bort från jorden i konstellationen Jungfrun. Arbetet utfördes på Uppsala universitet på institutionen för fysik och astronomi, eller Department of Physics and Astronomy. Detta gjordes huvudsakligen med hjälp av transmissionsspektroskopi och cross-correlation - en av de ledande metoderna idag för att analysera exoplanetära atmosfärer. Datan som använts för denna studie samlades in under en transit som skedde i mars 2022 med hjälp av spektrografen CRIRES+, stationerad vid Very Large Telescope (VLT) i Paranal. WASP-107 b klassas som en varm Jupiter, och eftersom de undersökta kemiska arterna huvudsakligen uppvisar spektrallinjer i det infraröda området (0.95-5.3 μm), är CRIRES+ ett sunt val då spektrografen är specialiserad på att undersöka infrarött ljus. Dataanalysen utfördes genom användningen av flertal script, byggda i Python med påföljande datareduktion. De datareduktionsmetoder som användes i detta syfte var ESO CRIRES+ standard data reduction pipeline, vilken inkluderar avlägsnandet av systematiska källor till brus såsom döda pixlar och den kosmiska bakgrundsstrålningen, och SysRem, vilket är en algoritm som används för att ta bort trender samt konstanta drag beroende på tid utmed varje pixelserie. I nuläget är SysRem en av de mer effektiva sätten att göra detta på, och är en vanlig metod i studier som denna. I detta projekt blev en exoplanet undersökt och dess atmosfär granskad för att se ifall H2O och CO förekommer i den. De cross-correlation templates som användes bestod av en som använde en kombination av båda kemiska arter, tillika en som endast detekterade CO. De detection maps som genererats från cross-correlation analysen föreslog först en ickedetektion för alla kombinationer av SysRem iterationer och templates, förutom två, vilka uppvisade signalement som möjligtvis indikerade en detektion, men utan särskild stark säkerhet. Dessa resultat föreslog en möjlig detektion av CO i atmosfären, men för att säkerställa detta krävs vidare undersökning.

Master of Science / Department of Statistics / Juan Du / Space and time are often vital components of research data sets. Accounting for and utilizing the space and time information in statistical models become beneficial when the response variable in question is proved to have a space and time dependence. This work focuses on the modeling and analysis of crop yield over space and time. Specifically, two different yield data sets were used. The first yield and environmental data set was collected across selected counties in Kansas from yield performance tests conducted for multiple years. The second yield data set was a survey data set collected by USDA across the US from 1900-2009. The objectives of our study were to investigate crop yield trends in space and time, quantify the variability in yield explained by genetics and space-time (environment) factors, and study how spatio-temporal information could be incorporated and also utilized in modeling and forecasting yield. Based on the format of these data sets, trend of irrigated and dryland crops was analyzed by employing time series statistical techniques. Some traditional linear regressions and smoothing techniques are first used to obtain the yield function. These models were then improved by incorporating time and space information either as explanatory variables or as auto- or cross- correlations adjusted in the residual covariance structures. In addition, a multivariate time series modeling approach was conducted to demonstrate how the space and time correlation information can be utilized to model and forecast yield and related variables. The conclusion from this research clearly emphasizes the importance of space and time components of data sets in research analysis. That is partly because they can often adjust (make up) for those underlying variables and factor effects that are not measured or not well understood.

International Telemetering Conference Proceedings / October 29-November 02, 1990 / Riviera Hotel and Convention Center, Las Vegas, Nevada / Various PN codes for use in TTC spread-spectrum systems are considered. The evaluation is based on peak magnitudes and amplitude distributions of both the even and the odd autocorrelation and crosscorrelation functions. Furthermore the influence of the phase of a sequence on the correlation parameters is studied, multiple-access characteristics in terms of the total interference parameter are evaluated and synchronous as well as asynchronous code generation is considered.1

International Telemetering Conference Proceedings / October 25-28, 1999 / Riviera Hotel and Convention Center, Las Vegas, Nevada / As testing requirements on the ranges require ever more sophisticated cross correlation of data from multiple data acquisition sources, it becomes increasingly advantageous to collect and disseminate this information in a more network oriented fashion. This allows any of the data collected at physically separated sites to be used simultaneously at multiple mission control or data reduction centers. This paper presents an approach that maximizes the use of legacy communication paths and data reduction systems to support an evolutionary migration toward the day when testing can take full advantage of commercial communication protocols and equipment such as OC-3, ATM, etc. One key element of this approach is the packetizing of data at each reception point to provide virtual circuit switching using packet routing. Based on the newly adopted IRIG/RCC 107-98 standard, the system may even be expanded all the way back to the actual sensors. The second key element is the use of the readily available time and timing pulses based on GPS to establish a uniform sampling interval that will allow the cross correlation of data received at different points spread over a wide area.

The accurate measurement of multiphase flows of oil/water/gas is a critical element of oil exploration and production. Thus, over the last three decades; the development and deployment of in-line multiphase flow metering systems has been a major focus worldwide. Accurate measurement of multiphase flow in the oil and gas industry is difficult because there is a wide range of flow regimes and multiphase meters do not generally perform well under the intermittent slug flow conditions which commonly occur in oil production. This thesis investigates the use of Doppler and cross-correlation ultrasonic measurements made in different high gas void fraction flow, partially separated liquid and gas flows, and homogeneous flow and raw slug flow, to assess the accuracy of measurement in these regimes. This approach has been tested on water/air flows in a 50mm diameter pipe facility. The system employs a partial gas/liquid separation and homogenisation using a T-Y junction configuration. A combination of ultrasonic measurement techniques was used to measure flow velocities and conductivity rings to measure the gas fraction. In the partially separated regime, ultrasonic cross-correlation and conductivity rings are used to measure the liquid flow-rate. In the homogeneous flow, a clamp-on ultrasonic Doppler meter is used to measure the homogeneous velocity and combined with conductivity ring measurements to provide measurement of the liquid and gas flow-rates. The slug flow regime measurements employ the raw Doppler shift data from the ultrasonic Doppler flowmeter, together with the slug flow closure equation and combined with gas fraction obtained by conductivity rings, to determine the liquid and gas flow-rates. Measurements were made with liquid velocities from 1.0m/s to 2.0m/s with gas void fractions up to 60%. Using these techniques the accuracies of the liquid flow-rate measurement in the partially separated, homogeneous and slug regimes were 10%, 10% and 15% respectively. The accuracy of the gas flow-rate in both the homogeneous and raw slug regimes was 10%. The method offers the possibility of further improvement in the accuracy by combining measurement from different regimes.

Maintenance of stable arterial blood pressure during orthostatic challenges is a major problem after spinal cord injury (SCI). Since early participation in rehabilitation is critically important in reducing long term morbidity, recovering the ability to regulate blood pressure during therapy is essential for individuals with SCI. The objective of our study was to investigate short term cardiovascular function of able-bodied (AB), paraplegic (PARA) and tetraplegic (TETRA) subjects in response to head up tilt (HUT) as an early indicator of autonomic damage that might forewarn of future orthostatic regulatory problems. We acquired cardiovascular variables from able-bodied (AB; n=11), paraplegic (PARA; n=5) and tetraplegic (TETRA; n=5) subjects in response to HUT. The SCI patients in both groups were in their first two months post injury. Data were recorded at rest and during 7 min each at 20??, 40??, 60?? and 80?? HUT. Techniques used to estimate regulatory capability and reflex activity included: Mean values and spectral power of heart rate (HR) and arterial blood pressure (BP), baroreflex sequence measurements and cross correlation between HR and systolic blood pressure (SBP). An index of baroreflex sensitivity (BRS), baroreflex effectiveness index (BEI), and the percentage occurrence of systolic blood pressure (BP) ramps and baroreflex sequences were calculated from baroreflex sequence measurements. The spectral power of HR and BP, the cross correlation of systolic BP and heart rate (HR) were examined in low frequency (LF: 0.04-0.15 Hz) and high frequency (HF: 0.15-0.4 Hz) ranges. The BRS index was significantly (p andlt; 0.05) decreased from supine to 80o HUT in AB and TETRA. This index in PARA was the lowest at each tilt position in the three groups, and decreased with tilt. The percentage of heart beats involved in systolic BP ramps and in baroreflex sequences significantly (pandlt;0.05) rose from supine to 80o HUT in AB, was relatively unchanged in PARA and declined in TETRA. Both of these indexes were significantly (pandlt;0.05) lowerin the SCI than in the AB group at each tilt level. The BEI values were greatest in AB, and declined with tilt in all groups. Spinal cord injured patients had less power of BP and HR fluctuations than AB in both LF and HF regions. The LF spectral power of BP and HR increased with tilt in AB, remained unchanged in PARA and decreased in TETRA. The HF spectral power of HR decreased in all three groups. The peak HR / BP cross correlation in the LF region was greatest in AB, and significantly (pandlt;0.05) increased during HUT in AB, remained fairly constant in PARA, and declined in TETRA. The peak cross correlation in the HF region significantly (pandlt;0.05) decreased with tilt in all groups, and the SCI group had lower values than AB at each tilt level. We conclude that both PARA and TETRA had a smaller percentage of SBP ramps, BRS, and lower BEI than AB, likely indicating decreased stimulation of arterial baroreceptors, and less engagement of feedback control. The mixed sympathetic, parasympathetic innervations of paraplegics, or their elevated HR, may contribute to their significantly lower BRS. Our data indicate that the pathways utilized to evoke baroreflex regulation of HR are compromised by SCI and this loss may be a major contributor to the decrease in orthostatic tolerance following injury.