Oregon Graduate Institute Research Articles

Remembering the Oregon Graduate Institute

Abstract The Oregon Graduate Center−later named the Oregon Graduate Institute−played a significant role in metallurgy in the Northwest for nearly four decades. This article reviews important work done by the organization in the second half of the twentieth century.

Language and Gender Classification of Speech Files Using Supervised Machine Learning Methods

ABSTRACTMany language identification (LID) systems are based on language models using techniques that consider the fluctuation of speech over time. Considering these fluctuations necessitates longer recording intervals to obtain reasonable accuracy. Our research extracts features from short recording intervals to enable successful classification of spoken language. The feature extraction process is based on frames of 20 ms, whereas most previous LIDs presented results based on much longer frames (3 s or longer). We defined and implemented 200 features divided into four feature sets: cepstrum features, RASTA features, spectrum features, and waveform features. We applied eight machine learning (ML) methods on the features that were extracted from a corpus containing speech files in 10 languages from the Oregon Graduate Institute (OGI) telephone speech database and compared their performances using extensive experimental evaluation. The best optimized classification results were achieved by random forest (RF): from 76.29% on 10 languages to 89.18% on 2 languages. These results are better or comparable to the state-of-the-art results for the OGI database. Another set of experiments that was performed was gender classification from 2 to 10 languages. The accuracy and the F measure values for the RF method for all the language experiments were greater than or equal to 90.05%.

Parametric representation of excitation source information for language identification

In this work, the linear prediction (LP) residual signal has been parameterized to capture the excitation source information for language identification (LID) study. LP residual signal has been processed at three different levels: sub-segmental, segmental and supra-segmental levels to demonstrate different aspects of language-specific excitation source information. Proposed excitation source features have been evaluated on 27 Indian languages from Indian Institute of Technology Kharagpur-Multi Lingual Indian Language Speech Corpus (IITKGP-MLILSC), Oregon Graduate Institute Multi-Language Telephone-based Speech (OGI-MLTS) and National Institute of Standards and Technology Language Recognition Evaluation (NIST LRE) 2011 corpora. LID systems were developed using Gaussian mixture model (GMM) and i-vector based approaches. Experimental results have shown that segmental level parametric features provide better identification accuracy (62%), compared to sub-segmental (40%) and supra-segmental level (34%) features. Excitation source features obtained from three levels show distinct language-specific evidence. Therefore, the scores from all three levels are combined to obtain the complete excitation source information for the LID task. LID performances achieved from both the excitation source and vocal tract system are compared. Finally, the scores obtained by processing the vocal tract and excitation source features are combined to achieve better improvement in LID accuracy. The best recognition accuracies obtained from stage-IV integrated LID systems I, II and III are 69%, 70% and 72% respectively.

Identification of Indian languages using multi-level spectral and prosodic features

In this paper spectral and prosodic features extracted from different levels are explored for analyzing the language specific information present in speech. In this work, spectral features extracted from frames of 20 ms (block processing), individual pitch cycles (pitch synchronous analysis) and glottal closure regions are used for discriminating the languages. Prosodic features extracted from syllable, tri-syllable and multi-word (phrase) levels are proposed in addition to spectral features for capturing the language specific information. In this study, language specific prosody is represented by intonation, rhythm and stress features at syllable and tri-syllable (words) levels, whereas temporal variations in fundamental frequency (F 0 contour), durations of syllables and temporal variations in intensities (energy contour) are used to represent the prosody at multi-word (phrase) level. For analyzing the language specific information in the proposed features, Indian language speech database (IITKGP-MLILSC) is used. Gaussian mixture models are used to capture the language specific information from the proposed features. The evaluation results indicate that language identification performance is improved with combination of features. Performance of proposed features is also analyzed on standard Oregon Graduate Institute Multi-Language Telephone-based Speech (OGI-MLTS) database.

Pitch synchronous and glottal closure based speech analysis for language recognition

This paper explores pitch synchronous and glottal closure (GC) based spectral features for analyzing the language specific information present in speech. For determining pitch cycles (for pitch synchronous analysis) and GC regions, instants of significant excitation (ISE) are used. The ISE correspond to the instants of glottal closure (epochs) in the case of voiced speech, and some random excitations like onset of burst in the case of nonvoiced speech. For analyzing the language specific information in the proposed features, Indian language speech database (IITKGP-MLILSC) is used. Gaussian mixture models are used to capture the language specific information from the proposed features. Proposed pitch synchronous and glottal closure spectral features are evaluated using language recognition studies. The evaluation results indicate that language recognition performance is better with pitch synchronous and GC based spectral features compared to conventional spectral features derived through block processing. GC based spectral features are found to be more robust against degradations due to background noise. Performance of proposed features is also analyzed on standard Oregon Graduate Institute Multi-Language Telephone-based Speech (OGI-MLTS) database.

Modest PGC-1α Overexpression in Muscle in Vivo Is Sufficient to Increase Insulin Sensitivity and Palmitate Oxidation in Subsarcolemmal, Not Intermyofibrillar, Mitochondria

PGC-1alpha overexpression in skeletal muscle, in vivo, has yielded disappointing and unexpected effects, including disrupted cellular integrity and insulin resistance. These unanticipated results may stem from an excessive PGC-1alpha overexpression in transgenic animals. Therefore, we examined the effects of a modest PGC-1alpha overexpression in a single rat muscle, in vivo, on fuel-handling proteins and insulin sensitivity. We also examined whether modest PGC-1alpha overexpression selectively targeted subsarcolemmal (SS) mitochondrial proteins and fatty acid oxidation, because SS mitochondria are metabolically more plastic than intermyofibrillar (IMF) mitochondria. Among metabolically heterogeneous rat hindlimb muscles, PGC-1alpha was highly correlated with their oxidative fiber content and with substrate transport proteins (GLUT4, FABPpm, and FAT/CD36) and mitochondrial proteins (COXIV and mTFA) but not with insulin-signaling proteins (phosphatidylinositol 3-kinase, IRS-1, and Akt2), nor with 5'-AMP-activated protein kinase, alpha2 subunit, and HSL. Transfection of PGC-1alpha into the red (RTA) and white tibialis anterior (WTA) compartments of the tibialis anterior muscle increased PGC-1alpha protein by 23-25%. This also induced the up-regulation of transport proteins (FAT/CD36, 35-195%; GLUT4, 20-32%) and 5'-AMP-activated protein kinase, alpha2 subunit (37-48%), but not other proteins (FABPpm, IRS-1, phosphatidylinositol 3-kinase, Akt2, and HSL). SS and IMF mitochondrial proteins were also up-regulated, including COXIV (15-75%), FAT/CD36 (17-30%), and mTFA (15-85%). PGC-1alpha overexpression also increased palmitate oxidation in SS (RTA, +116%; WTA, +40%) but not in IMF mitochondria, and increased insulin-stimulated phosphorylation of AKT2 (28-43%) and rates of glucose transport (RTA, +20%; WTA, +38%). Thus, in skeletal muscle in vivo, a modest PGC-1alpha overexpression up-regulated selected plasmalemmal and mitochondrial fuel-handling proteins, increased SS (not IMF) mitochondrial fatty acid oxidation, and improved insulin sensitivity.

Bimakalim: A Promising KATP Channel Activating Agent

Laboratory of Cardiovascular Pharmacology, Department of Cardiac Surgery and II Section of Cardiology, University “La Sapienza”; Rome, Italy; Department of Chemistry, Biochemistry, and Molecular Biology, Oregon Graduate Institute of Science and Technology, Portland, OR USA; on leave from Laboratory of Experimental Anesthesiology and Cellular Physiology, University, Caen, France, in partial fulfillment of work performed during the tenure of a Doctorate exchange fellowship between the Universities of Poitiers, France and Rome “La Sapienza,” Italy

Language identification using acoustic log-likelihoods of syllable-like units

Automatic spoken language identification (LID) is the task of identifying the language from a short utterance of the speech signal uttered by an unknown speaker. The most successful approach to LID uses phone recognizers of several languages in parallel [Zissman, M.A., 1996. Comparison of four approaches to automatic language identification of telephone speech. IEEE Trans. Speech Audio Process. 4 (1), 31–44]. The basic requirement to build a parallel phone recognition (PPR) system is segmented and labeled speech corpora. In this paper, a novel approach is proposed for the LID task which uses parallel syllable-like unit recognizers, in a frame work similar to the PPR approach in the literature. The difference is that the sub-word unit models for each of the languages to be recognized are generated in an unsupervised manner without the use of segmented and labeled speech corpora. The training data of each of the languages is first segmented into syllable-like units and language-dependent syllable-like unit inventory is created. These syllable-like units are then clustered using an incremental approach. This results in a set of syllable-like units models for each language. Using these language-dependent syllable-like unit models, language identification is performed based on accumulated acoustic log-likelihoods. Our initial results on the Oregon Graduate Institute Multi-language Telephone Speech Corpus [Muthusamy, Y.K., Cole, R.A., Oshika, B.T., 1992. The OGI multi-language telephone speech corpus. In: Proceedings of Internat. Conf. Spoken Language Process., October 1992, pp. 895–898] show that the performance is 72.3%. We further show that if only a subset of syllable-like unit models that are unique (in some sense) are considered, the performance improves to 75.9%.

A Critical Assessment of Recent Work on the Application of Gas/Particle Partitioning Theories to Cigarette Smoke

Abstract In the August 1997 issue of Environmental Science&Technology (ES&T), PANKOW and co-workers at the Oregon Graduate Institute reported that the addition of ammonia-containing additives to cigarette tobacco increased the amount of unprotonated nicotine in cigarette mainstream smoke (MSS) and thus increased the bioavailability of nicotine to the smoker. Articles about PANKOW's work also appeared in other publications along with allegations that ammonia-containing additives are used to manipulate nicotine deliveries. However, initial review of PANKOW's research and that reported on environmental tobacco smoke (ETS) in an earlier paper showed that potentially serious issues existed with PANKOW's experimental data and conclusions. Consequently, a critical assessment of PANKOW's research and the underlying theories of gas/particle partitioning was undertaken. This assessment confirmed that PANKOW and his co-workers made a number of errors not only in their determinations of the gas/particle partitioning coefficients for nicotine in MSS and ETS but also in the interpretations of the data. During the preparation of this assessment, data from other researchers became public. These data showed that there was no correlation between tobacco ammonia (including residual ammonia from the use of ammonia-containing additives) and MSS ammonia deliveries and MSS smoke pH, and that the amount of unprotonated nicotine in the undiluted MSS of a full flavor (FF) American filter cigarette was less than 0.1%. These new data fully substantiated the findings of this assessment, and it can be safely concluded that the assertions made by PANKOW and his co-workers were incorrect. However, this assessment also showed that there is significant merit in the application of PANKOW's theory of absorptive partitioning for the estimation of the gas/particle partitioning of semivolatile components in MSS and ETS. Application of PANKOW's theory along with data from recent tobacco related conferences has allowed estimation of the gas/particle partition coefficient for nicotine in cigarette MSS and also has allowed approximation of values for the activity coefficient of nicotine in mainstream particulate matter.

Data-Driven Temporal Filters and Alternatives to GMM in Speaker Verification

Malayath, Narendranath, Hermansky, Hynek, Kajarekar, Sachin, and Yegnanarayana, B., Data-Driven Temporal Filters and Alternatives to GMM in Speaker Verification, Digital Signal Processing10(2000), 55–74.This paper discusses the research directions pursued jointly at the Anthropic Signal Processing Group of the Oregon Graduate Institute and at the Speech and Vision Laboratory of the Indian Institute of Technology Madras. Current methods for speaker verification are based on modeling the speaker characteristics using Gaussian mixture models (GMM). The performance of these systems significantly degrades if the target speakers use a telephone handset that is different from that used while training. Conventional methods for channel normalization include utterance-based mean subtraction (MS) and RelAtive SpecTrAl (RASTA) filtering. In this paper we introduce a novel method for designing filters that are capable of normalizing the variability introduced by different telephone handsets. The design of the filter is based on the estimated second-order statistics of handset variability. This filter is applied on the logarithmic energy outputs of Mel spaced filter banks. We also demonstrate the effectiveness of the proposed channel normalizing filter in improving speaker verification performance in mismatched conditions. GMM-based systems often use thousands of mixture components and hence require a large number of parameters to characterize each target speaker. In order to address this issue we propose an alternative to GMM for modeling speaker characteristics. The alternative is based on speaker-specific mapping and it relies on a speaker-independent representation of speech.

Operating systems specialization (invited talk) (abstract only)

This talk will present a systems-centric view of the opportunities, effectiveness, and challenges of applying specialization techniques to operating systems. The first part of the talk will focus on our experiences using manual and then tool-assisted specialization in the systematic optimization of legacy operating systems software. These experiences were gained in the Synthetix project which explored the use of static, dynamic and incremental optimistic specialization in a variety of operating system contexts, including file system calls in HP-UX, Linux signal delivery, Berkeley packet filter processing, and Sun RPC. The second part of the talk will focus on the lessons learned in Synthetix, the implications for systems, and the key opportunities and challenges for future research. Jonathan Walpole is a Professor in the Computer Science and Engineering Department at Oregon Graduate Institute. His research interests are in the area of adaptive systems software and its application in distributed, mobile, multimedia computing environments. His work has focused on quality of service specification, adaptive resource management and dynamic specialization for enhanced performance, survivability and evolvability of large software systems.

Ten myths of multimodal interaction

article Free Access Share on Ten myths of multimodal interaction Author: Sharon Oviatt Oregon Graduate Institute of Science and Technology, Beaverton Oregon Graduate Institute of Science and Technology, BeavertonView Profile Authors Info & Claims Communications of the ACMVolume 42Issue 11Nov. 1999 pp 74–81https://doi.org/10.1145/319382.319398Published:01 November 1999Publication History 365citation7,987DownloadsMetricsTotal Citations365Total Downloads7,987Last 12 Months446Last 6 weeks38 Get Citation AlertsNew Citation Alert added!This alert has been successfully added and will be sent to:You will be notified whenever a record that you have chosen has been cited.To manage your alert preferences, click on the button below.Manage my Alerts New Citation Alert!Please log in to your account Save to BinderSave to BinderCreate a New BinderNameCancelCreateExport CitationPublisher SiteView all FormatsPDF

Imprecise Exceptions, Co-Inductively

In a recent paper, Peyton Jones et al. proposed a design for imprecise exceptions in the lazy functional programming language Haskell [PJRH+99]. The main contribution of the design was that it allowed the language to continue to enjoy its current rich algebra of transformations. However, the denotational semantics used to formalise the design does not combine easily with other extensions, most notably that of concurrency. We present an alternative semantics for a lazy functional language with imprecise exceptions which is entirely operational in nature, and combines well with other extensions, such as I/O and concurrency. The semantics is based upon a convergence relation, which describes evaluation, and an exceptional convergence relation, which describes the raising of exceptions. Convergence and exceptional convergence lead naturally to a simple notion of refinement, where a term M is refined by N whenever they have identical convergent behaviour, and any exception raised by N can also be raised by M. We are able to validate many call-by-name equivalences and standard program transformations, including the ubiquitous strictness transformation.The first author gratefully acknowledges the hospitality of the Microsoft Research Lab (Cambridge), where this work was begun, and the Oregon Graduate Institute, where this work was completed. The second author was supported by grant number GR/L38356 from the U.K. EPSRC.

Peroxo-, Oxo-, and Hydroxo-Bridged Dicopper Complexes: Observation of Exogenous Hydrocarbon Substrate Oxidation

ADVERTISEMENT RETURN TO ISSUEPREVCommunicationNEXTPeroxo-, Oxo-, and Hydroxo-Bridged Dicopper Complexes: Observation of Exogenous Hydrocarbon Substrate OxidationHonorio V. Obias, Yin Lin, N. N. Murthy, Elna Pidcock, Edward I. Solomon, Martina Ralle, Ninian J. Blackburn, Yorck-Michael Neuhold, Andreas D. Zuberbühler, and Kenneth D. KarlinView Author Information Department of Chemistry The Johns Hopkins University Baltimore, Maryland 21218 Department of Chemistry, Stanford University Stanford, California 94305 Department of Chemistry, Biochemistry, and Molecular Biology, Oregon Graduate Institute of Science & Technology, Portland, Oregon 97291 Institute für Anorganische Chemie University of Basel, CH-4056 Basel, Switzerland Cite this: J. Am. Chem. Soc. 1998, 120, 49, 12960–12961Publication Date (Web):November 26, 1998Publication History Received29 May 1998Published online26 November 1998Published inissue 1 December 1998https://doi.org/10.1021/ja981878hCopyright © 1998 American Chemical SocietyRIGHTS & PERMISSIONSArticle Views1736Altmetric-Citations138LEARN ABOUT THESE METRICSArticle Views are the COUNTER-compliant sum of full text article downloads since November 2008 (both PDF and HTML) across all institutions and individuals. These metrics are regularly updated to reflect usage leading up to the last few days.Citations are the number of other articles citing this article, calculated by Crossref and updated daily. Find more information about Crossref citation counts.The Altmetric Attention Score is a quantitative measure of the attention that a research article has received online. Clicking on the donut icon will load a page at altmetric.com with additional details about the score and the social media presence for the given article. Find more information on the Altmetric Attention Score and how the score is calculated. Share Add toView InAdd Full Text with ReferenceAdd Description ExportRISCitationCitation and abstractCitation and referencesMore Options Share onFacebookTwitterWechatLinked InReddit Read OnlinePDF (50 KB) Get e-AlertsSupporting Info (1)»Supporting Information Supporting Information SUBJECTS:Chemical structure,Ligands,Reactivity,Resonance structures,Solvents Get e-Alerts

The Microsoft database research group

Microsoft’s strategic interest in the database field dates from 1993 and the efforts of David Vaskevitch, who is now the Microsoft Vice President in charge of the database and transaction processing product development groups. David’s vision was that the world would need millions of servers, and that this presented a wonderful opportunity to a company like Microsoft that sells software in high volume and at low prices. Database systems played an important role in Vaskevitch’s vision, and, indeed, in Microsoft’s current product plans. David began looking for premier database and transaction processing people in late 1993. The scope of Vaskevitch’s efforts included a desire for Microsoft to establish a database research group. Rick Rashid, Microsoft Research Vice President, collaborated with Vaskevitch in recruiting David Lomet from Digital’s Cambridge Research Lab to initiate the Microsoft Database Research Group. Lomet joined Microsoft Research in January of 1995. Hence, Microsoft’s Database Research Group is now a little over three and a half years old. One person does not a group make. Recruiting efforts continued. Surajit Chaudhuri, a researcher from HP Labs joined the Database Group in February of 1996. Paul Larson, a professor from the University of Waterloo joined in May of that year. Vivek Narasayya was initially an intern as a graduate student from the University of Washington in the summer of 1996, officially joining the group in April of 1997. Roger Barga, the newest member of the group and a new Oregon Graduate Institute Ph.D., joined in December, 1997.

Response to Comment on “Redox Interactions of Cr(Vi) and Substituted Phenols: Products and Mechanism”

ADVERTISEMENT RETURN TO ISSUEPREVCorrespondence/Rebut...Correspondence/RebuttalNEXTResponse to Comment on “Redox Interactions of Cr(Vi) and Substituted Phenols: Products and Mechanism”William Fish and Michael S. ElovitzView Author Information Department of Environmental Science and Engineering Oregon Graduate Institute of Science & Technology 20000 NW Walker Road P.O. Box 91000 Portland, Oregon 97291-1000Cite this: Environ. Sci. Technol. 1998, 32, 15, 2345Publication Date (Web):June 4, 1998Publication History Published online4 June 1998Published inissue 1 August 1998https://doi.org/10.1021/es972019gCopyright © 1998 American Chemical SocietyRIGHTS & PERMISSIONSArticle Views150Altmetric-Citations-LEARN ABOUT THESE METRICSArticle Views are the COUNTER-compliant sum of full text article downloads since November 2008 (both PDF and HTML) across all institutions and individuals. These metrics are regularly updated to reflect usage leading up to the last few days.Citations are the number of other articles citing this article, calculated by Crossref and updated daily. Find more information about Crossref citation counts.The Altmetric Attention Score is a quantitative measure of the attention that a research article has received online. Clicking on the donut icon will load a page at altmetric.com with additional details about the score and the social media presence for the given article. Find more information on the Altmetric Attention Score and how the score is calculated. Share Add toView InAdd Full Text with ReferenceAdd Description ExportRISCitationCitation and abstractCitation and referencesMore Options Share onFacebookTwitterWechatLinked InReddit PDF (17 KB) Get e-AlertsSUBJECTS:Absorption,Anions,Diodes,Organic compounds,Spectroscopy Get e-Alerts

Intelligent animated agents for interactive language training

This report describes a three-year project, now eight months old, to develop interactive learning tools for language training with profoundly deaf children. The tools combine four key technologies: speech recognition, developed at the Oregon Graduate Institute; speech synthesis, developed at the University of Edinburgh and modified at OGI; facial animation, developed at University of California, Santa Cruz; and face tracking and speech reading, developed at Carnegie Mellon University. These technologies are being combined to create an intelligent conversational agent; a three-dimensional face that produces and understands auditory and visual speech. The agent has been incorporated into CSLU Toolkit, a software environment for developing and researching spoken language systems. We describe our experiences in bringing interactive learning tools to classrooms at the Tucker-Maxon Oral School in Portland, Oregon, and the technological advances that are required for this project to succeed.

Integrating recent research results into undergraduate curricula (panel)

On July 7-11, 60 computer scientists came together at The Evergreen State College in Olympia, Washington to consider how the undergraduate curricula might be improved in light of recent research in computer science. At this NSF-funded workshop, researchers presented work in four areas where current research might be particularly relevant, and undergraduate faculty (with experience in software engineering, functional programming, artificial intelligence, discrete mathematics or theoretical computer science) explored the current state of undergraduate computer science curricula and ways in which they might be better informed by recent research.Each participant attended sessions in one of the four interest areas--Software Engineering Capstone Courses, Functional Programming, Neural Networks and Their Applications, and Computational Geometry--and faculty developed curricular materials that they could use in their teaching the following year. Those materials are being placed on the WWW, and faculty are refining them as they use them in their courses. A second workshop is planned for summer, 1998.The workshop was sponsored by The Evergreen State College, the Oregon Graduate Institute, the Washington Center for the Improvement of Undergraduate Education, and by the Center for Discrete Mathematics and Theoretical Computer Science (DIMACS).All of the panelists participated in the workshop and each panel member will share his or her experiences in incorporating the recent research results studied at the workshop into their undergraduate curriculum. In addition, the panelists will discuss with the audience their own plans for integrating research results into their own undergraduate programs.

Research on Domain-specific Embedded Languages and Program Generators

Embedding is the process of implementing a language by defining functions in an existing “host” Language; the host language with these added functions is the new language. As a consequence; the new language comes equipped with all the features of the host language, with no additional work on the part of the language designer. Embedding works particularly well when the host language is a functional language.We describe several examples of embedded languages. The first is a language for specifying simple pictures. The others are program generators, that is, languages used to specify programs in other languages. In all of these examples, the host language is Standard ML; in the program generating languages, the target language is C+plus;. The power obtained from the host language is the main emphasis of our presentation.The author gratefully acknowledges the support provided by the Oregon Graduate Institute, where he was on sabbatical during the preparation of this paper.

The Student Watershed Research Project (SWRP): collecting watershed data for use by resource management agencies

From 1992 to 1994, Saturday Academy and the Oregon Graduate Institute of Science and Technology instituted the pilot phase of a long-term monitoring program called the Student Watershed Research Project (SWRP). The SWRP program was developed to create and maintain collaboration among eighth through twelfth grade teachers and students, scientists, businesses, governmental agencies, and community members which would couple environmental education with the collection of high quality, reproducible watershed health data.