Skip to main content

John Gowdy

Rensselaer Polytechnic Institute, Economics, Faculty Member

Followers

65

Following

9

Co-authors

2

Public Views

I am a retired economics professor. My areas of interest are evolutionary theory, social evolution, inequality and power, the human threats to non-human nature, ultrasociality (ants, termites, and humans with agriculture), the agricultural transition, hunter-gatherer societies, climate change, uncivilization.

less

Noel B. Salazar

KU Leuven

Kristen J Gremillion

Ohio State University

Armando Marques-Guedes

UNL - New University of Lisbon

Università di Bologna

Yannis Hamilakis

Brown University

Field Museum

The University of Newcastle

Santiago Ginnobili

Universidad de Buenos Aires

Johns Hopkins University

Richard Cloutier

Université du Québec à Rimouski

Interests

Uploads

Papers by John Gowdy

Hexachlorophane--yes or no?

BMJ, Nov 19, 1977

Age and death in breast cancer

BMJ, Jul 21, 1979

Subband based classification of speech under stress

Proceedings of the 1998 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP '98 (Cat. No.98CH36181)

This study proposes a new set of feature parameters based on subband analysis of the speech signa... more This study proposes a new set of feature parameters based on subband analysis of the speech signal for classi cation of speech under stress. The new speech features are Scale Energy SE, Autocorrelation-Scale-Energy ACSE, Subband based cepstral parameters SC, and Autocorrelation-SC ACSC. The parameters' ability to capture di erent stress types is compared to widely used Mel-scale cepstrum based representations: Mel-frequency cepstral coe cents MFCC and Autocorrelation-Mel-scale AC-Mel. Next, a feedforward neural network is formulated for speaker-dependent stress classi cation of 10 stress conditions: Angry, Clear, Cond50 70, Fast, Loud, Lombard, Neutral, Question, Slow, and Soft. The classi cation algorithm is evaluated using a previously established stressed speech database SUSAS 4. Subband based features are shown to achieve + 7:3 and +9:1 increase in the classi cation rates over the MFCC based parameters for ungrouped and grouped stress closed vocabulary test scenarios respectively. Moreover the average scores across the simulations of new features are +8:6 and +13:6 higher than MFCC based features for the ungrouped and grouped stress test scenarious respectively.

Myoelectric signals for multimodal speech recognition

Interspeech 2005, 2005

A Coupled Hidden Markov Model (CHMM) is proposed in this paper to perform multimodal speech recog... more A Coupled Hidden Markov Model (CHMM) is proposed in this paper to perform multimodal speech recognition using myoeletric signals (MES) from the muscles of vocal articulation. MES signals are immune to noise, and words that are acoustically similar manifest distinctly in MES. Hence, they would effectively complement the acoustic data in a multimodal speech recognition system. Research in AudioVisual Speech Recognition has shown that CHMMs model the asynchrony between different data streams effectively. Hence, we propose CHMM for multimodal speech recognition using audio and MES as the two data streams. Our experiments indicate that the multimodal CHMM system significantly outperforms the audio only system at different SNRs. We have also provided a comparison between different features for MES and have found that wavelet features provide the best results.

Stereo-based elliptical head tracking

2005 13th European Signal Processing Conference, 2005

A novel algorithm for head-tracking based on stereo is presented in this paper. Employing stereo ... more A novel algorithm for head-tracking based on stereo is presented in this paper. Employing stereo vision makes the tracker robust to many factors such as clutter, color and lighting variations, that affect contemporary head trackers. The head is modeled as an ellipse and tracking is performed on the foreground obtained by modeling depth. This is facilitated by using stereo vision. Further, we present a simple head size estimation technique that is critical for modeling considerable motion by the subject. The results obtained demonstrate the robustness of the head tracker and the accuracy of the head-size estimation technique. This would find extensive use in speaker tracking in smart rooms.

Combination of Linear Prediction and Phase Decomposition for Glottal Source Analysis on Voiced Speech

ArXiv, 2016

Some glottal analysis approaches based upon linear prediction or complex cepstrum approaches have... more Some glottal analysis approaches based upon linear prediction or complex cepstrum approaches have been proved to be effective to estimate glottal source from real speech utterances. We propose a new approach employing both an all-pole odd-order linear prediction to provide a coarse estimation and phase decomposition based causality/anti-causality separation to generate further refinements. The obtained measures show that this method improved performance in terms of reducing source-filter separation in estimation of glottal flow pulses (GFP). No glottal model fitting is required by this method, thus it has wide and flexible adaptation to retain fidelity of speakers's vocal features with computationally affordable resource. The method is evaluated on real speech utterances to validate it.

Language modeling using Independent Component Analysis for Automatic Speech Recognition

2005 13th European Signal Processing Conference, 2005

Conventional statistical language models such as N-grams are inadequate to model long distance de... more Conventional statistical language models such as N-grams are inadequate to model long distance dependencies in natural language. In this paper we propose a novel statistical language model to capture topic related long range dependencies. Humans have the inherent ability to identify long range dependencies in natural language. Given a set of related words humans can easily identify the context in which the set of words is occurring. It has been shown by many researchers that Independent Component Analysis (ICA) captures these kind of dependencies better than any other formulation. Furthermore, ICA provides a topic decomposition that can be easily interpreted by humans compared to other models. This paper describes the development of a language model using ICA. The topic model is combined with a standard N-gram to produce the language model. The perplexity results obtained show that this language model is a viable language model for speech recognition purposes.

An audio-visual approach to simultaneous-speaker speech recognition

2003 IEEE International Conference on Acoustics, Speech, and Signal Processing, 2003. Proceedings. (ICASSP '03).

Audiovisual speech recognition is an area with great potential to help solve challenging problems... more Audiovisual speech recognition is an area with great potential to help solve challenging problems in speech processing. Difficulties due to background noises are significantly reduced by the additional information provided by extra visual features. The presence of additional speech from other talkers during recording may be viewed as one of the most difficult sources of noise. This paper presents a study using audiovisual speech recognition for simultaneous-speaker speech recognition. The desired goal is to separate and potentially recognize speech from several simultaneous speakers. Speaker pairs from the CUAVE multimodal speech corpus are used in this work. Audiovisual techniques are compared against speaker-independent and speaker-dependent audio-only methods for speech recognition of individuals from these pairs. For information on obtaining CUAVE, please visit the following web page (http://ece.clemson.edu/speech).

Real time eye tracking for human computer interfaces

2003 International Conference on Multimedia and Expo. ICME '03. Proceedings (Cat. No.03TH8698), 2003

In recent years considerable interest has developed in real time eye tracking for various applica... more In recent years considerable interest has developed in real time eye tracking for various applications. An approach that has received a lot of attention is the use of infrared technology for purposes of eye tracking. In this paper we propose a technique that does not rely on the use of infrared devices for eye tracking. Instead, our eye tracker makes use of a binary classifier with a dynamic training strategy and an unsupervised clustering stage in order to efficiently track the pupil (eyeball) in real time. The dynamic training strategy makes the algorithm subject (speaker) and lighting condition invariant. Our algorithm does not make any assumption regarding the position of the speaker's face in the field of view of the camera, nor does it restrict the 'natural' motion of the speaker in the field of view of the camera. Experimental results from a real time implementation show that this algorithm is robust and able to detect the pupils under various illumination conditions.

Development and Use of Simulation Modules for Teaching a Distance-Learning Course on Digital Processing of Speech Signals

This paper describes the development, utilization, and effectiveness of a set of interactive simu... more This paper describes the development, utilization, and effectiveness of a set of interactive simulation modules used with a distance-learning version of a graduate class on Digital Processing of Speech Signals at Clemson University. The motivation for developing these modules was to give distance-learning students a special learning tool, to partially compensate for their disadvantages of being separated from campus laboratories, the course instructor, and other students. Although the modules were developed primarily for students taking the course remotely, they have also enhanced the learning experiences of on-campus students. The paper describes the following simulation modules: phoneme segmentation, speech production model, speech coding, speech synthesis using LPC, speech recognition, and speech analysis. The authors have also developed a system for homework distribution, submission, and grading over the Web. This includes JAVA-based tools that permit students to generate and submit diagrams as part of their homework solutions.

Multi-platform CBI tools using Linux and Java-based solutions for distance learning

Acoustics, Speech, and Signal Processing, 1988. ICASSP-88., 1988 International Conference on

This paper presents a set of software tools developed to enhance local and distance learning for ... more This paper presents a set of software tools developed to enhance local and distance learning for speech and signal processing classes. The multi-platform paradigm is stressed in both computer-based instruction (CBI) and World Wide Web (WWW), Java-based tools. The CBI speech tools are a package developed to enhance a graduate course in digital speech processing; the tools may be run on various platforms and will benefit local or distance learners taking the course via teleconferencing. Students may gain “hands-on” experience using this package on their own personal computer (PC) or a university computer. The WWW, Java-based tools have been designed to be used for interactive homework in signal processing classes but also have a wide range of applicability. JavaGram is the name for the currently developed application that allows students to easily “turn-in” homework diagrams, such as signal-flow charts, via the WWW. This is an improvement over past systems that use only forms in WWW h...

Modified multiband excitation model at 2400 bps

1997 IEEE International Conference on Acoustics, Speech, and Signal Processing

This paper presents the Modied Multiband Excitation Model used for speech coding. In many MBE mod... more This paper presents the Modied Multiband Excitation Model used for speech coding. In many MBE model coders, speech quality is degraded when incorrect voicing decisions are made, particularly for high-pitched female speakers. The MMBE addresses this issue with a modied voiced/unvoiced decision algorithm and a more robust pitch estimate. The listening quality o f speech produced using the MMBE model is superior to the FS-1016 CELP coder and is at least comparable with the new 2400 bps MELP coder chosen as the new 2400 bps Federal Standard.

Concurrent renal-cell carcinoma and cutaneous leiomyomas: A case of HLRCC

Radiology Case Reports, 2015

A 51-year-old Caucasian female initially presented to the emergency department in April 2010 with... more A 51-year-old Caucasian female initially presented to the emergency department in April 2010 with complaints of urinary retention. She had a past medical history significant for lupus, hypertension, stage III chronic renal failure, multiple painful skin lesions (Figs. 1, 2), and severe uterine fibroids requiring hysterectomy at age 23. Ultrasound (US) and noncontrast computed tomography (CT) evaluation showed a cortically based, partially exophytic mass, involving the lower pole of the right kidney, that demonstrated peripheral calcifications concerning for neoplasm (Fig. 3). A magnetic resonance imaging (MRI) scan performed one month later revealed two enhancing lesions in the right kidney that were concerning for neoplasm, one of which corresponded to the partially calcified lesion seen on CT. The patient underwent partial nephrectomy for the right renal lesions in October 2010. Pathological exam revealed

DBN based multi-stream models for audio-visual speech recognition

2004 IEEE International Conference on Acoustics, Speech, and Signal Processing

In this paper, we propose a model based on Dynamic Bayesian Networks (DBNs) to integrate informat... more In this paper, we propose a model based on Dynamic Bayesian Networks (DBNs) to integrate information from multiple audio and visual streams. We also compare the DBN based system (implemented using the Graphical Model Toolkit (GMTK)) with a classical HMM (implemented in the Hidden Markov Model Toolkit (HTK)) for both the single and two stream integration problems. We also propose a new model (mixed integration) to integrate information from three or more streams derived from different modalities and compare the new model's performance with that of a synchronous integration scheme. A new technique to estimate stream confidence measures for the integration of three or more streams is also developed and implemented. Results from our implementation using the Clemson University Audio Visual Experiments (CUAVE) database indicate an absolute improvement of about £ ¥ ¤ in word accuracy in the-4 to 10db average case when making use of two audio and one video streams for the mixed integration models over the sychronous models.

Improving the Recognition of Eating Gestures Using Intergesture Sequential Dependencies

IEEE Journal of Biomedical and Health Informatics, 2015

This paper considers the problem of recognizing eating gestures by tracking wrist motion. Eating ... more This paper considers the problem of recognizing eating gestures by tracking wrist motion. Eating gestures are activities commonly undertaken during the consumption of a meal, such as sipping a drink of liquid or using utensils to cut food. Each of these gestures causes a pattern of wrist motion that can be tracked to automatically identify the activity. Previous works have studied this problem at the level of a single gesture. In this paper, we demonstrate that individual gestures have sequential dependence. To study this, three types of classifiers were built: 1) a K-nearest neighbor classifier which uses no sequential context, 2) a hidden Markov model (HMM) which captures the sequential context of subgesture motions, and 3) HMMs that model intergesture sequential dependencies. We built first-order to sixth-order HMMs to evaluate the usefulness of increasing amounts of sequential dependence to aid recognition. On a dataset of 25 meals, we found that the baseline accuracies for the KNN and the subgesture HMM classifiers were 75.8% and 84.3%, respectively. Using HMMs that model intergesture sequential dependencies, we were able to increase accuracy to up to 96.5%. These results demonstrate that sequential dependencies exist between eating gestures and that they can be exploited to improve recognition accuracy.

Multi-stream product modal audio-visual integration strategy for robust adaptive speech recognition

IEEE International Conference on Acoustics Speech and Signal Processing, 2002

In this paper, we extend an existing audio-only automatic speech recognizer to implement a multi-... more In this paper, we extend an existing audio-only automatic speech recognizer to implement a multi-stream audiovisual automatic speech recognition (AV-ASR) system. Our method forms a multi-stream feature vector from audiovisual speech data, computes the statistical modal parameters probabilities on the basis of multi-stream audiovisual features, and performs dynamic programming jointly on the multi-stream product modal Hidden Markov Models (MS-PM-HMMs) by utilizing a noise type and signal-to-noise ratio (SNR) based stream-weighting value. Experimental results are presented for an isolated word recognition task for eight different noise types from the NOISEX data base for several SNR values. The proposed system reduces the word error rate (WER), averaged over several SNR and noise types, from 55.9% with the audio-only recognizer and 7.9% with the late-integration audiovisual recognizer to 2.6% WER in the validation set.

CUAVE: a new audio-visual database for multimodal human-computer interface research

IEEE International Conference on Acoustics Speech and Signal Processing, 2002

Applied mel-frequency discrete wavelet coefficients and parallel model compensation for noise-robust speech recognition

Speech Communication, 2006

Interfering noise severely degrades the performance of a speech recognition system. The Parallel ... more Interfering noise severely degrades the performance of a speech recognition system. The Parallel Model Compensation (PMC) technique is one of the most efficient techniques for dealing with such noise. Another approach is to use features local in the frequency domain, such as Mel-Frequency Discrete Wavelet Coefficients (MFDWCs). In this paper, we investigate the use of PMC and MFDWC features to take advantage of both noise compensation and local features (MFDWCs) to decrease the effect of noise on recognition performance. We also introduce a practical weighting technique based on the noise level of each coefficient. We evaluate the performance of several wavelet-schemes using the NOISEX-92 database for various noise types and noise levels. Finally, we compare the performance of these versus Mel-Frequency Cepstral Coefficients (MFCCs), both using PMC. Experimental results show significant performance improvements for MFDWCs versus MFCCs, particularly after compensating the HMMs using the PMC technique. The best feature vector among the six MFDWCs we tried gave 13.72 and 5.29 points performance improvement, on the average, over MFCCs for À6 and 0 dB SNR, respectively. This corresponds to 39.9% and 62.8% error reductions, respectively. Weighting the partial score of each coefficient based on the noise level further improves the performance. The average error rates for the best MFDWCs dropped from 19.57% to 16.71% and from 3.14% to 2.14% for À6 dB and 0 dB noise levels, respectively, using the weighting scheme. These improvements correspond to 14.6% and 31.8% error reductions for À6 dB and 0 dB noise levels, respectively.

Investigating the Possibility of Improving Recognition Robustness with Applied Chaos Theory

One of the primary difficulties involved with speech recognition is dealing with sounds other tha... more One of the primary difficulties involved with speech recognition is dealing with sounds other than speech that may occur while a person is speaking. There have been many techniques that attempt to deal with background "noise." Many of these make use of some type of stochastic model to decrease the effect of the interfering signal. This paper, however, investigates the possibility of a new technique involving nonlinear chaos theory. In the past few decades, the idea of chaos has been of growing interest in physics and has recently begun to enter the world of signal processing with some good results. In a simple sense, a chaotic signal is one that appears random but has some underlying, nearly deterministic structure. This paper considers using some of the knowledge from chaos theory in modeling background "noise" that may interfere with a recognizer or other speech system.

Application of affine-invariant Fourier descriptors to recognition of 3-D objects

IEEE Transactions on Pattern Analysis and Machine Intelligence, 1990

In this work, the method of Fourier descriptors has been extended to produce a set of normalized ... more In this work, the method of Fourier descriptors has been extended to produce a set of normalized coefficients which are invariant under any affine transformation (translation, rotation, scaling, and shearing). The method is based on a parameterized boundary description which is transformed to the Fourier domain and normalized there to eliminate dependencies on the affine transformation and on the starting point. Invariance to affine transforms allows considerable robustness when applied to images of objects which rotate in all three dimensions. This is demonstrated by processing silhouettes of aircraft as the aircraft maneuver in three-space.

Hexachlorophane--yes or no?

BMJ, Nov 19, 1977

Age and death in breast cancer

BMJ, Jul 21, 1979

Subband based classification of speech under stress

Proceedings of the 1998 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP '98 (Cat. No.98CH36181)

This study proposes a new set of feature parameters based on subband analysis of the speech signa... more This study proposes a new set of feature parameters based on subband analysis of the speech signal for classi cation of speech under stress. The new speech features are Scale Energy SE, Autocorrelation-Scale-Energy ACSE, Subband based cepstral parameters SC, and Autocorrelation-SC ACSC. The parameters' ability to capture di erent stress types is compared to widely used Mel-scale cepstrum based representations: Mel-frequency cepstral coe cents MFCC and Autocorrelation-Mel-scale AC-Mel. Next, a feedforward neural network is formulated for speaker-dependent stress classi cation of 10 stress conditions: Angry, Clear, Cond50 70, Fast, Loud, Lombard, Neutral, Question, Slow, and Soft. The classi cation algorithm is evaluated using a previously established stressed speech database SUSAS 4. Subband based features are shown to achieve + 7:3 and +9:1 increase in the classi cation rates over the MFCC based parameters for ungrouped and grouped stress closed vocabulary test scenarios respectively. Moreover the average scores across the simulations of new features are +8:6 and +13:6 higher than MFCC based features for the ungrouped and grouped stress test scenarious respectively.

Myoelectric signals for multimodal speech recognition

Interspeech 2005, 2005

A Coupled Hidden Markov Model (CHMM) is proposed in this paper to perform multimodal speech recog... more A Coupled Hidden Markov Model (CHMM) is proposed in this paper to perform multimodal speech recognition using myoeletric signals (MES) from the muscles of vocal articulation. MES signals are immune to noise, and words that are acoustically similar manifest distinctly in MES. Hence, they would effectively complement the acoustic data in a multimodal speech recognition system. Research in AudioVisual Speech Recognition has shown that CHMMs model the asynchrony between different data streams effectively. Hence, we propose CHMM for multimodal speech recognition using audio and MES as the two data streams. Our experiments indicate that the multimodal CHMM system significantly outperforms the audio only system at different SNRs. We have also provided a comparison between different features for MES and have found that wavelet features provide the best results.

Stereo-based elliptical head tracking

2005 13th European Signal Processing Conference, 2005

A novel algorithm for head-tracking based on stereo is presented in this paper. Employing stereo ... more A novel algorithm for head-tracking based on stereo is presented in this paper. Employing stereo vision makes the tracker robust to many factors such as clutter, color and lighting variations, that affect contemporary head trackers. The head is modeled as an ellipse and tracking is performed on the foreground obtained by modeling depth. This is facilitated by using stereo vision. Further, we present a simple head size estimation technique that is critical for modeling considerable motion by the subject. The results obtained demonstrate the robustness of the head tracker and the accuracy of the head-size estimation technique. This would find extensive use in speaker tracking in smart rooms.

Combination of Linear Prediction and Phase Decomposition for Glottal Source Analysis on Voiced Speech

ArXiv, 2016

Some glottal analysis approaches based upon linear prediction or complex cepstrum approaches have... more Some glottal analysis approaches based upon linear prediction or complex cepstrum approaches have been proved to be effective to estimate glottal source from real speech utterances. We propose a new approach employing both an all-pole odd-order linear prediction to provide a coarse estimation and phase decomposition based causality/anti-causality separation to generate further refinements. The obtained measures show that this method improved performance in terms of reducing source-filter separation in estimation of glottal flow pulses (GFP). No glottal model fitting is required by this method, thus it has wide and flexible adaptation to retain fidelity of speakers's vocal features with computationally affordable resource. The method is evaluated on real speech utterances to validate it.

Language modeling using Independent Component Analysis for Automatic Speech Recognition

2005 13th European Signal Processing Conference, 2005

Conventional statistical language models such as N-grams are inadequate to model long distance de... more Conventional statistical language models such as N-grams are inadequate to model long distance dependencies in natural language. In this paper we propose a novel statistical language model to capture topic related long range dependencies. Humans have the inherent ability to identify long range dependencies in natural language. Given a set of related words humans can easily identify the context in which the set of words is occurring. It has been shown by many researchers that Independent Component Analysis (ICA) captures these kind of dependencies better than any other formulation. Furthermore, ICA provides a topic decomposition that can be easily interpreted by humans compared to other models. This paper describes the development of a language model using ICA. The topic model is combined with a standard N-gram to produce the language model. The perplexity results obtained show that this language model is a viable language model for speech recognition purposes.

An audio-visual approach to simultaneous-speaker speech recognition

2003 IEEE International Conference on Acoustics, Speech, and Signal Processing, 2003. Proceedings. (ICASSP '03).

Audiovisual speech recognition is an area with great potential to help solve challenging problems... more Audiovisual speech recognition is an area with great potential to help solve challenging problems in speech processing. Difficulties due to background noises are significantly reduced by the additional information provided by extra visual features. The presence of additional speech from other talkers during recording may be viewed as one of the most difficult sources of noise. This paper presents a study using audiovisual speech recognition for simultaneous-speaker speech recognition. The desired goal is to separate and potentially recognize speech from several simultaneous speakers. Speaker pairs from the CUAVE multimodal speech corpus are used in this work. Audiovisual techniques are compared against speaker-independent and speaker-dependent audio-only methods for speech recognition of individuals from these pairs. For information on obtaining CUAVE, please visit the following web page (http://ece.clemson.edu/speech).

Real time eye tracking for human computer interfaces

2003 International Conference on Multimedia and Expo. ICME '03. Proceedings (Cat. No.03TH8698), 2003

In recent years considerable interest has developed in real time eye tracking for various applica... more In recent years considerable interest has developed in real time eye tracking for various applications. An approach that has received a lot of attention is the use of infrared technology for purposes of eye tracking. In this paper we propose a technique that does not rely on the use of infrared devices for eye tracking. Instead, our eye tracker makes use of a binary classifier with a dynamic training strategy and an unsupervised clustering stage in order to efficiently track the pupil (eyeball) in real time. The dynamic training strategy makes the algorithm subject (speaker) and lighting condition invariant. Our algorithm does not make any assumption regarding the position of the speaker's face in the field of view of the camera, nor does it restrict the 'natural' motion of the speaker in the field of view of the camera. Experimental results from a real time implementation show that this algorithm is robust and able to detect the pupils under various illumination conditions.

Development and Use of Simulation Modules for Teaching a Distance-Learning Course on Digital Processing of Speech Signals

This paper describes the development, utilization, and effectiveness of a set of interactive simu... more This paper describes the development, utilization, and effectiveness of a set of interactive simulation modules used with a distance-learning version of a graduate class on Digital Processing of Speech Signals at Clemson University. The motivation for developing these modules was to give distance-learning students a special learning tool, to partially compensate for their disadvantages of being separated from campus laboratories, the course instructor, and other students. Although the modules were developed primarily for students taking the course remotely, they have also enhanced the learning experiences of on-campus students. The paper describes the following simulation modules: phoneme segmentation, speech production model, speech coding, speech synthesis using LPC, speech recognition, and speech analysis. The authors have also developed a system for homework distribution, submission, and grading over the Web. This includes JAVA-based tools that permit students to generate and submit diagrams as part of their homework solutions.

Multi-platform CBI tools using Linux and Java-based solutions for distance learning

Acoustics, Speech, and Signal Processing, 1988. ICASSP-88., 1988 International Conference on

This paper presents a set of software tools developed to enhance local and distance learning for ... more This paper presents a set of software tools developed to enhance local and distance learning for speech and signal processing classes. The multi-platform paradigm is stressed in both computer-based instruction (CBI) and World Wide Web (WWW), Java-based tools. The CBI speech tools are a package developed to enhance a graduate course in digital speech processing; the tools may be run on various platforms and will benefit local or distance learners taking the course via teleconferencing. Students may gain “hands-on” experience using this package on their own personal computer (PC) or a university computer. The WWW, Java-based tools have been designed to be used for interactive homework in signal processing classes but also have a wide range of applicability. JavaGram is the name for the currently developed application that allows students to easily “turn-in” homework diagrams, such as signal-flow charts, via the WWW. This is an improvement over past systems that use only forms in WWW h...

Modified multiband excitation model at 2400 bps

1997 IEEE International Conference on Acoustics, Speech, and Signal Processing

This paper presents the Modied Multiband Excitation Model used for speech coding. In many MBE mod... more This paper presents the Modied Multiband Excitation Model used for speech coding. In many MBE model coders, speech quality is degraded when incorrect voicing decisions are made, particularly for high-pitched female speakers. The MMBE addresses this issue with a modied voiced/unvoiced decision algorithm and a more robust pitch estimate. The listening quality o f speech produced using the MMBE model is superior to the FS-1016 CELP coder and is at least comparable with the new 2400 bps MELP coder chosen as the new 2400 bps Federal Standard.

Concurrent renal-cell carcinoma and cutaneous leiomyomas: A case of HLRCC

Radiology Case Reports, 2015

A 51-year-old Caucasian female initially presented to the emergency department in April 2010 with... more A 51-year-old Caucasian female initially presented to the emergency department in April 2010 with complaints of urinary retention. She had a past medical history significant for lupus, hypertension, stage III chronic renal failure, multiple painful skin lesions (Figs. 1, 2), and severe uterine fibroids requiring hysterectomy at age 23. Ultrasound (US) and noncontrast computed tomography (CT) evaluation showed a cortically based, partially exophytic mass, involving the lower pole of the right kidney, that demonstrated peripheral calcifications concerning for neoplasm (Fig. 3). A magnetic resonance imaging (MRI) scan performed one month later revealed two enhancing lesions in the right kidney that were concerning for neoplasm, one of which corresponded to the partially calcified lesion seen on CT. The patient underwent partial nephrectomy for the right renal lesions in October 2010. Pathological exam revealed

DBN based multi-stream models for audio-visual speech recognition

2004 IEEE International Conference on Acoustics, Speech, and Signal Processing

In this paper, we propose a model based on Dynamic Bayesian Networks (DBNs) to integrate informat... more In this paper, we propose a model based on Dynamic Bayesian Networks (DBNs) to integrate information from multiple audio and visual streams. We also compare the DBN based system (implemented using the Graphical Model Toolkit (GMTK)) with a classical HMM (implemented in the Hidden Markov Model Toolkit (HTK)) for both the single and two stream integration problems. We also propose a new model (mixed integration) to integrate information from three or more streams derived from different modalities and compare the new model's performance with that of a synchronous integration scheme. A new technique to estimate stream confidence measures for the integration of three or more streams is also developed and implemented. Results from our implementation using the Clemson University Audio Visual Experiments (CUAVE) database indicate an absolute improvement of about £ ¥ ¤ in word accuracy in the-4 to 10db average case when making use of two audio and one video streams for the mixed integration models over the sychronous models.

Improving the Recognition of Eating Gestures Using Intergesture Sequential Dependencies

IEEE Journal of Biomedical and Health Informatics, 2015

This paper considers the problem of recognizing eating gestures by tracking wrist motion. Eating ... more This paper considers the problem of recognizing eating gestures by tracking wrist motion. Eating gestures are activities commonly undertaken during the consumption of a meal, such as sipping a drink of liquid or using utensils to cut food. Each of these gestures causes a pattern of wrist motion that can be tracked to automatically identify the activity. Previous works have studied this problem at the level of a single gesture. In this paper, we demonstrate that individual gestures have sequential dependence. To study this, three types of classifiers were built: 1) a K-nearest neighbor classifier which uses no sequential context, 2) a hidden Markov model (HMM) which captures the sequential context of subgesture motions, and 3) HMMs that model intergesture sequential dependencies. We built first-order to sixth-order HMMs to evaluate the usefulness of increasing amounts of sequential dependence to aid recognition. On a dataset of 25 meals, we found that the baseline accuracies for the KNN and the subgesture HMM classifiers were 75.8% and 84.3%, respectively. Using HMMs that model intergesture sequential dependencies, we were able to increase accuracy to up to 96.5%. These results demonstrate that sequential dependencies exist between eating gestures and that they can be exploited to improve recognition accuracy.

Multi-stream product modal audio-visual integration strategy for robust adaptive speech recognition

IEEE International Conference on Acoustics Speech and Signal Processing, 2002

In this paper, we extend an existing audio-only automatic speech recognizer to implement a multi-... more In this paper, we extend an existing audio-only automatic speech recognizer to implement a multi-stream audiovisual automatic speech recognition (AV-ASR) system. Our method forms a multi-stream feature vector from audiovisual speech data, computes the statistical modal parameters probabilities on the basis of multi-stream audiovisual features, and performs dynamic programming jointly on the multi-stream product modal Hidden Markov Models (MS-PM-HMMs) by utilizing a noise type and signal-to-noise ratio (SNR) based stream-weighting value. Experimental results are presented for an isolated word recognition task for eight different noise types from the NOISEX data base for several SNR values. The proposed system reduces the word error rate (WER), averaged over several SNR and noise types, from 55.9% with the audio-only recognizer and 7.9% with the late-integration audiovisual recognizer to 2.6% WER in the validation set.

CUAVE: a new audio-visual database for multimodal human-computer interface research

IEEE International Conference on Acoustics Speech and Signal Processing, 2002

Applied mel-frequency discrete wavelet coefficients and parallel model compensation for noise-robust speech recognition

Speech Communication, 2006

Interfering noise severely degrades the performance of a speech recognition system. The Parallel ... more Interfering noise severely degrades the performance of a speech recognition system. The Parallel Model Compensation (PMC) technique is one of the most efficient techniques for dealing with such noise. Another approach is to use features local in the frequency domain, such as Mel-Frequency Discrete Wavelet Coefficients (MFDWCs). In this paper, we investigate the use of PMC and MFDWC features to take advantage of both noise compensation and local features (MFDWCs) to decrease the effect of noise on recognition performance. We also introduce a practical weighting technique based on the noise level of each coefficient. We evaluate the performance of several wavelet-schemes using the NOISEX-92 database for various noise types and noise levels. Finally, we compare the performance of these versus Mel-Frequency Cepstral Coefficients (MFCCs), both using PMC. Experimental results show significant performance improvements for MFDWCs versus MFCCs, particularly after compensating the HMMs using the PMC technique. The best feature vector among the six MFDWCs we tried gave 13.72 and 5.29 points performance improvement, on the average, over MFCCs for À6 and 0 dB SNR, respectively. This corresponds to 39.9% and 62.8% error reductions, respectively. Weighting the partial score of each coefficient based on the noise level further improves the performance. The average error rates for the best MFDWCs dropped from 19.57% to 16.71% and from 3.14% to 2.14% for À6 dB and 0 dB noise levels, respectively, using the weighting scheme. These improvements correspond to 14.6% and 31.8% error reductions for À6 dB and 0 dB noise levels, respectively.

Investigating the Possibility of Improving Recognition Robustness with Applied Chaos Theory

One of the primary difficulties involved with speech recognition is dealing with sounds other tha... more One of the primary difficulties involved with speech recognition is dealing with sounds other than speech that may occur while a person is speaking. There have been many techniques that attempt to deal with background "noise." Many of these make use of some type of stochastic model to decrease the effect of the interfering signal. This paper, however, investigates the possibility of a new technique involving nonlinear chaos theory. In the past few decades, the idea of chaos has been of growing interest in physics and has recently begun to enter the world of signal processing with some good results. In a simple sense, a chaotic signal is one that appears random but has some underlying, nearly deterministic structure. This paper considers using some of the knowledge from chaos theory in modeling background "noise" that may interfere with a recognizer or other speech system.

Application of affine-invariant Fourier descriptors to recognition of 3-D objects

IEEE Transactions on Pattern Analysis and Machine Intelligence, 1990

In this work, the method of Fourier descriptors has been extended to produce a set of normalized ... more In this work, the method of Fourier descriptors has been extended to produce a set of normalized coefficients which are invariant under any affine transformation (translation, rotation, scaling, and shearing). The method is based on a parameterized boundary description which is transformed to the Fourier domain and normalized there to eliminate dependencies on the affine transformation and on the starting point. Invariance to affine transforms allows considerable robustness when applied to images of objects which rotate in all three dimensions. This is demonstrated by processing silhouettes of aircraft as the aircraft maneuver in three-space.