Study on the Effects of Intrinsic Variation using i-vectors in Text-Independent Speaker Verification

Size: px
Start display at page:

Download "Study on the Effects of Intrinsic Variation using i-vectors in Text-Independent Speaker Verification"

Transcription

1 Study on the Effects of Intrinsic Variation using i-vectors in Text-Independent Speaker Verification Sheng Chen, Mingxing Xu, Emlyn Pratt Department of Computer Science and Technology Tsinghua University, Beijing, China June 26, /19

2 Outline 1 Introduction Main Challenges Problem and Proposal 2 i-vector Framework for Intrinsic Variation Modeling i-vector Framework Compensation for Intrinsic Variability 3 Experiments Data partitions for train and testing Description of Speaker Verification systems Experimental Results 4 Conclusions & Future Work 2/19

3 Main Challenges in Speaker Verification Extrinsic Variability Mismatched Channels Environmental Noise... Intrinsic Variability Speaking Style Emotion Speech Volume State of Health... 3/19

4 Problem and Proposal Problem we focus on: Performances of speaker verification systems are adversely affected by intrinsic variability. Question: How the speaker verification system perform when enrollment and testing are done in mismatched conditions due to intrinsic variability? How the technologies focused on modeling the total variability behave in addressing the effects of intrinsic variability in speaker verification? Proposal: Model the intrinsic variability with i-vector framework. 4/19

5 How to define variation forms? Reading Speaking Style English Speaking Language Angry Happy Emotional State Neutral Spontaneous speech at normal rate and volume in Chinese Speaking Rate Fast Slow Speaking Volume Loud Soft Whispered Physical Status Mumbled Denasalized 5/19

6 i-vector Framework for Intrinsic Variation Modeling Application of i-vector modeling: Effective for Channel Compensation 1 i-vector Framework Total Variability M = m+tw Cosine Similarity Scoring score(w target,w test ) = wtarget,wtest w target w test Idea: How about modeling the Intrinsic Variability with i-vector Framework? Front-end factor analysis for speaker verification, N Dehak, PJ Kenny, R Dehak, 6/19

7 How to remove the effects of intrinsic variations? Linear Discriminant Analysis(LDA) Idea: Minimize the within-speaker variability while maximizing the between-speaker variability S B v = λs W v Within-Class Covariance Normalization(WCCN) Idea: Deemphasize the direction of high intra-speaker variability W 1 = BB t Nuisance Attribute Projection(NAP) Idea: Remove the nuisance direction P = I VV t 7/19

8 Experiments Experimental Data Intrinsic Variation Corpus Data partitions for train and testing Description of Speaker Verification Systems GMM-UBM baseline system i-vector based speaker verification systems Experimental Results 8/19

9 Intrinsic Variation Corpus Type Description Number of variation forms 12 Number of Subjects 110(46 males, 64 females) Format WAVE Duration 180s Sample Rate 8KHz Resolution 8 bits Soundtrack Mono 9/19

10 Data partitions in the intrinsic variation corpus Function Source Description UBM traing data Training data used for total variability space Training data used for LDA,WCCN and NAP Testing data 30 speakers 30 speakers 20 speakers 20 speakers 18 hours 12 variation forms 18 hours 12 variation forms 12 hours 12 variation forms 2400 utterances 12 variation forms 10/19

11 Description of Speaker Verification systems GMM-UBM (Baseline System) P(x λ) = M ω i g(x,µ i,σ i ) i=1 S(U) = logp(u λ TAR ) logp(u λ UBM ) Feature: 39 dimensional MFCC UBM: 512 Gaussian mixtures i-vector based Speaker Verification Systems i-vector + LDA i-vector + WCCN i-vector + NAP i-vector + LDA + WCCN 200 dimensional i-vector 11/19

12 EERs(%) for each enrollment condition when testing utterances contain the twelve variation forms Speech Variation Variation Form GMM-UBM LDA WCCN NAP LDA+WCCN Base Case Spontaneous Speaking Style Reading Speaking Volume Speaking Rate Emotional State Physical Status Loud Soft Whispered Fast Slow Angry Happy Denasalized Mumbled Speaking Language English /19

13 Performances of i-vector based systems Overall EER(%) of Speaker Verification systems in the intrinsic variation corpus System EER(%) Relative Reduction(%) GMM-UBM(baseline) i-vector+lda i-vector+wccn i-vector+nap i-vector+lda+wccn /19

14 DET curve for GMM-UBM based system and four i-vector based systems. Speaker Detection Performance 60 GMM-UBM i-vector+lda i-vector+nap i-vector+wccn i-vector+lda+wccn 40 Miss probability (in %) False Alarm probability (in %) 14/19

15 Comparison between GMM-UBM and i-vector in matched and mismatched conditions 15/19

16 EERs(%) for each testing condition when spontaneous utterances are used for enrollment Speech Variation Variation Form GMM-UBM LDA WCCN NAP LDA+WCCN Base Case Spontaneous Speaking Style Reading Speaking Volume Speaking Rate Emotional State Physical Status Loud Soft Whispered Fast Slow Angry Happy Denasalized Mumbled Speaking Language English /19

17 EERs(%) for each testing condition when whispering utterances are used for enrollment Speech Variation Variation Form GMM-UBM LDA WCCN NAP LDA+WCCN Base Case Spontaneous Speaking Style Reading Speaking Volume Speaking Rate Emotional State Physical Status Loud Soft Whispered Fast Slow Angry Happy Denasalized Mumbled Speaking Language English /19

18 Conclusions & Future Work Conclusions Mismatches in intrinsic variations cause sharp degradation in speaker verification performance. The i-vector framwork performs better than GMM-UBM in modeling intrinsic variations. Whispering utterances bring the largest degradation of speaker verification performances. Future Work More techniques for intrinsic variation compensation. Improvements in feature domain. 18/19

19 Q & A Thanks! 19/19

IEEE Proof. Web Version. PROGRESSIVE speaker adaptation has been considered

IEEE Proof. Web Version. PROGRESSIVE speaker adaptation has been considered IEEE TRANSACTIONS ON AUDIO, SPEECH, AND LANGUAGE PROCESSING 1 A Joint Factor Analysis Approach to Progressive Model Adaptation in Text-Independent Speaker Verification Shou-Chun Yin, Richard Rose, Senior

More information

EFFECTS OF BACKGROUND DATA DURATION ON SPEAKER VERIFICATION PERFORMANCE

EFFECTS OF BACKGROUND DATA DURATION ON SPEAKER VERIFICATION PERFORMANCE Uludağ Üniversitesi Mühendislik-Mimarlık Fakültesi Dergisi, Cilt 18, Sayı 1, 2013 ARAŞTIRMA EFFECTS OF BACKGROUND DATA DURATION ON SPEAKER VERIFICATION PERFORMANCE Cemal HANİLÇİ * Figen ERTAŞ * Abstract:

More information

Deep Neural Network Approaches to Speaker and Language Recognition

Deep Neural Network Approaches to Speaker and Language Recognition IEEE SIGNAL PROCESSING LETTERS, VOL. 22, NO. 10, OCTOBER 2015 1671 Deep Neural Network Approaches to Speaker and Language Recognition Fred Richardson, Senior Member, IEEE, Douglas Reynolds, Fellow, IEEE,

More information

The effect of mismatched recording conditions on human and automatic speaker recognition in forensic applications

The effect of mismatched recording conditions on human and automatic speaker recognition in forensic applications Forensic Science International 146S (2004) S95 S99 www.elsevier.com/locate/forsciint The effect of mismatched recording conditions on human and automatic speaker recognition in forensic applications A.

More information

ARMORVOX IMPOSTORMAPS HOW TO BUILD AN EFFECTIVE VOICE BIOMETRIC SOLUTION IN THREE EASY STEPS

ARMORVOX IMPOSTORMAPS HOW TO BUILD AN EFFECTIVE VOICE BIOMETRIC SOLUTION IN THREE EASY STEPS ARMORVOX IMPOSTORMAPS HOW TO BUILD AN EFFECTIVE VOICE BIOMETRIC SOLUTION IN THREE EASY STEPS ImpostorMaps is a methodology developed by Auraya and available from Auraya resellers worldwide to configure,

More information

ABC System description for NIST SRE 2010

ABC System description for NIST SRE 2010 ABC System description for NIST SRE 2010 May 6, 2010 1 Introduction The ABC submission is a collaboration between: Agnitio Labs, South Africa Brno University of Technology, Czech Republic CRIM, Canada

More information

Emotion Detection from Speech

Emotion Detection from Speech Emotion Detection from Speech 1. Introduction Although emotion detection from speech is a relatively new field of research, it has many potential applications. In human-computer or human-human interaction

More information

Functional Auditory Performance Indicators (FAPI)

Functional Auditory Performance Indicators (FAPI) Functional Performance Indicators (FAPI) An Integrated Approach to Skill FAPI Overview The Functional (FAPI) assesses the functional auditory skills of children with hearing loss. It can be used by parents,

More information

Automatic Evaluation Software for Contact Centre Agents voice Handling Performance

Automatic Evaluation Software for Contact Centre Agents voice Handling Performance International Journal of Scientific and Research Publications, Volume 5, Issue 1, January 2015 1 Automatic Evaluation Software for Contact Centre Agents voice Handling Performance K.K.A. Nipuni N. Perera,

More information

Training Universal Background Models for Speaker Recognition

Training Universal Background Models for Speaker Recognition Odyssey 2010 The Speaer and Language Recognition Worshop 28 June 1 July 2010, Brno, Czech Republic Training Universal Bacground Models for Speaer Recognition Mohamed Kamal Omar and Jason Pelecanos IBM

More information

ADAPTIVE AND DISCRIMINATIVE MODELING FOR IMPROVED MISPRONUNCIATION DETECTION. Horacio Franco, Luciana Ferrer, and Harry Bratt

ADAPTIVE AND DISCRIMINATIVE MODELING FOR IMPROVED MISPRONUNCIATION DETECTION. Horacio Franco, Luciana Ferrer, and Harry Bratt ADAPTIVE AND DISCRIMINATIVE MODELING FOR IMPROVED MISPRONUNCIATION DETECTION Horacio Franco, Luciana Ferrer, and Harry Bratt Speech Technology and Research Laboratory, SRI International, Menlo Park, CA

More information

Available from Deakin Research Online:

Available from Deakin Research Online: This is the authors final peered reviewed (post print) version of the item published as: Adibi,S 2014, A low overhead scaled equalized harmonic-based voice authentication system, Telematics and informatics,

More information

AS indicated by the growing number of participants in

AS indicated by the growing number of participants in 1960 IEEE TRANSACTIONS ON AUDIO, SPEECH, AND LANGUAGE PROCESSING, VOL. 15, NO. 7, SEPTEMBER 2007 State-of-the-Art Performance in Text-Independent Speaker Verification Through Open-Source Software Benoît

More information

APPLYING MFCC-BASED AUTOMATIC SPEAKER RECOGNITION TO GSM AND FORENSIC DATA

APPLYING MFCC-BASED AUTOMATIC SPEAKER RECOGNITION TO GSM AND FORENSIC DATA APPLYING MFCC-BASED AUTOMATIC SPEAKER RECOGNITION TO GSM AND FORENSIC DATA Tuija Niemi-Laitinen*, Juhani Saastamoinen**, Tomi Kinnunen**, Pasi Fränti** *Crime Laboratory, NBI, Finland **Dept. of Computer

More information

TranSegId: A System for Concurrent Speech Transcription, Speaker Segmentation and Speaker Identification

TranSegId: A System for Concurrent Speech Transcription, Speaker Segmentation and Speaker Identification TranSegId: A System for Concurrent Speech Transcription, Speaker Segmentation and Speaker Identification Mahesh Viswanathan, Homayoon S.M. Beigi, Alain Tritschler IBM Thomas J. Watson Research Labs Research

More information

Establishing the Uniqueness of the Human Voice for Security Applications

Establishing the Uniqueness of the Human Voice for Security Applications Proceedings of Student/Faculty Research Day, CSIS, Pace University, May 7th, 2004 Establishing the Uniqueness of the Human Voice for Security Applications Naresh P. Trilok, Sung-Hyuk Cha, and Charles C.

More information

Security in Voice Authentication

Security in Voice Authentication Security in Voice Authentication by Chenguang Yang A Dissertation Submitted to the Faculty of the WORCESTER POLYTECHNIC INSTITUTE In partial fulfillment of the requirements for the Degree of Doctor of

More information

How to Improve the Sound Quality of Your Microphone

How to Improve the Sound Quality of Your Microphone An Extension to the Sammon Mapping for the Robust Visualization of Speaker Dependencies Andreas Maier, Julian Exner, Stefan Steidl, Anton Batliner, Tino Haderlein, and Elmar Nöth Universität Erlangen-Nürnberg,

More information

Speech Signal Processing: An Overview

Speech Signal Processing: An Overview Speech Signal Processing: An Overview S. R. M. Prasanna Department of Electronics and Electrical Engineering Indian Institute of Technology Guwahati December, 2012 Prasanna (EMST Lab, EEE, IITG) Speech

More information

MACHINE LEARNING IN HIGH ENERGY PHYSICS

MACHINE LEARNING IN HIGH ENERGY PHYSICS MACHINE LEARNING IN HIGH ENERGY PHYSICS LECTURE #1 Alex Rogozhnikov, 2015 INTRO NOTES 4 days two lectures, two practice seminars every day this is introductory track to machine learning kaggle competition!

More information

On sequence kernels for SVM classification of sets of vectors: application to speaker verification

On sequence kernels for SVM classification of sets of vectors: application to speaker verification On sequence kernels for SVM classification of sets of vectors: application to speaker verification Major part of the Ph.D. work of In collaboration with Jérôme Louradour Francis Bach (ARMINES) within E-TEAM

More information

A CHINESE SPEECH DATA WAREHOUSE

A CHINESE SPEECH DATA WAREHOUSE A CHINESE SPEECH DATA WAREHOUSE LUK Wing-Pong, Robert and CHENG Chung-Keng Department of Computing, Hong Kong Polytechnic University Tel: 2766 5143, FAX: 2774 0842, E-mail: {csrluk,cskcheng}@comp.polyu.edu.hk

More information

Membering T M : A Conference Call Service with Speaker-Independent Name Dialing on AIN

Membering T M : A Conference Call Service with Speaker-Independent Name Dialing on AIN PAGE 30 Membering T M : A Conference Call Service with Speaker-Independent Name Dialing on AIN Sung-Joon Park, Kyung-Ae Jang, Jae-In Kim, Myoung-Wan Koo, Chu-Shik Jhon Service Development Laboratory, KT,

More information

Example: Credit card default, we may be more interested in predicting the probabilty of a default than classifying individuals as default or not.

Example: Credit card default, we may be more interested in predicting the probabilty of a default than classifying individuals as default or not. Statistical Learning: Chapter 4 Classification 4.1 Introduction Supervised learning with a categorical (Qualitative) response Notation: - Feature vector X, - qualitative response Y, taking values in C

More information

1816 IEEE TRANSACTIONS ON IMAGE PROCESSING, VOL. 15, NO. 7, JULY 2006. Principal Components Null Space Analysis for Image and Video Classification

1816 IEEE TRANSACTIONS ON IMAGE PROCESSING, VOL. 15, NO. 7, JULY 2006. Principal Components Null Space Analysis for Image and Video Classification 1816 IEEE TRANSACTIONS ON IMAGE PROCESSING, VOL. 15, NO. 7, JULY 2006 Principal Components Null Space Analysis for Image and Video Classification Namrata Vaswani, Member, IEEE, and Rama Chellappa, Fellow,

More information

SPEAKER IDENTIFICATION FROM YOUTUBE OBTAINED DATA

SPEAKER IDENTIFICATION FROM YOUTUBE OBTAINED DATA SPEAKER IDENTIFICATION FROM YOUTUBE OBTAINED DATA Nitesh Kumar Chaudhary 1 and Shraddha Srivastav 2 1 Department of Electronics & Communication Engineering, LNMIIT, Jaipur, India 2 Bharti School Of Telecommunication,

More information

Solutions to Exam in Speech Signal Processing EN2300

Solutions to Exam in Speech Signal Processing EN2300 Solutions to Exam in Speech Signal Processing EN23 Date: Thursday, Dec 2, 8: 3: Place: Allowed: Grades: Language: Solutions: Q34, Q36 Beta Math Handbook (or corresponding), calculator with empty memory.

More information

Voice Communication Package v7.0 of front-end voice processing software technologies General description and technical specification

Voice Communication Package v7.0 of front-end voice processing software technologies General description and technical specification Voice Communication Package v7.0 of front-end voice processing software technologies General description and technical specification (Revision 1.0, May 2012) General VCP information Voice Communication

More information

IBM Research Report. CSR: Speaker Recognition from Compressed VoIP Packet Stream

IBM Research Report. CSR: Speaker Recognition from Compressed VoIP Packet Stream RC23499 (W0501-090) January 19, 2005 Computer Science IBM Research Report CSR: Speaker Recognition from Compressed Packet Stream Charu Aggarwal, David Olshefski, Debanjan Saha, Zon-Yin Shae, Philip Yu

More information

Bi-Modal Person Recognition on a Mobile Phone: using mobile phone data

Bi-Modal Person Recognition on a Mobile Phone: using mobile phone data Bi-Modal Person Recognition on a Mobile Phone: using mobile phone data Chris McCool, Sébastien Marcel, Abdenour Hadid, Matti Pietikäinen, Pavel Matějka, Jan Černocký, Norman Poh, Josef Kittler, Anthony

More information

Automatic Cross-Biometric Footstep Database Labelling using Speaker Recognition

Automatic Cross-Biometric Footstep Database Labelling using Speaker Recognition Automatic Cross-Biometric Footstep Database Labelling using Speaker Recognition Ruben Vera-Rodriguez 1, John S.D. Mason 1 and Nicholas W.D. Evans 1,2 1 Speech and Image Research Group, Swansea University,

More information

Thirukkural - A Text-to-Speech Synthesis System

Thirukkural - A Text-to-Speech Synthesis System Thirukkural - A Text-to-Speech Synthesis System G. L. Jayavardhana Rama, A. G. Ramakrishnan, M Vijay Venkatesh, R. Murali Shankar Department of Electrical Engg, Indian Institute of Science, Bangalore 560012,

More information

Measuring Performance in a Biometrics Based Multi-Factor Authentication Dialog. A Nuance Education Paper

Measuring Performance in a Biometrics Based Multi-Factor Authentication Dialog. A Nuance Education Paper Measuring Performance in a Biometrics Based Multi-Factor Authentication Dialog A Nuance Education Paper 2009 Definition of Multi-Factor Authentication Dialog Many automated authentication applications

More information

Channel-dependent GMM and Multi-class Logistic Regression models for language recognition

Channel-dependent GMM and Multi-class Logistic Regression models for language recognition Channel-dependent GMM and Multi-class Logistic Regression models for language recognition David A. van Leeuwen TNO Human Factors Soesterberg, the Netherlands david.vanleeuwen@tno.nl Niko Brümmer Spescom

More information

SPEECH DATA MINING, SPEECH ANALYTICS, VOICE BIOMETRY. www.phonexia.com, 1/41

SPEECH DATA MINING, SPEECH ANALYTICS, VOICE BIOMETRY. www.phonexia.com, 1/41 SPEECH DATA MINING, SPEECH ANALYTICS, VOICE BIOMETRY www.phonexia.com, 1/41 OVERVIEW How to move speech technology from research labs to the market? What are the current challenges is speech recognition

More information

Turkish Radiology Dictation System

Turkish Radiology Dictation System Turkish Radiology Dictation System Ebru Arısoy, Levent M. Arslan Boaziçi University, Electrical and Electronic Engineering Department, 34342, Bebek, stanbul, Turkey arisoyeb@boun.edu.tr, arslanle@boun.edu.tr

More information

ROBUST TEXT-INDEPENDENT SPEAKER IDENTIFICATION USING SHORT TEST AND TRAINING SESSIONS

ROBUST TEXT-INDEPENDENT SPEAKER IDENTIFICATION USING SHORT TEST AND TRAINING SESSIONS 18th European Signal Processing Conference (EUSIPCO-21) Aalborg, Denmark, August 23-27, 21 ROBUST TEXT-INDEPENDENT SPEAKER IDENTIFICATION USING SHORT TEST AND TRAINING SESSIONS Christos Tzagkarakis and

More information

Secure-Access System via Fixed and Mobile Telephone Networks using Voice Biometrics

Secure-Access System via Fixed and Mobile Telephone Networks using Voice Biometrics Secure-Access System via Fixed and Mobile Telephone Networks using Voice Biometrics Anastasis Kounoudes 1, Anixi Antonakoudi 1, Vasilis Kekatos 2 1 The Philips College, Computing and Information Systems

More information

Hardware Implementation of Probabilistic State Machine for Word Recognition

Hardware Implementation of Probabilistic State Machine for Word Recognition IJECT Vo l. 4, Is s u e Sp l - 5, Ju l y - Se p t 2013 ISSN : 2230-7109 (Online) ISSN : 2230-9543 (Print) Hardware Implementation of Probabilistic State Machine for Word Recognition 1 Soorya Asokan, 2

More information

An Overview of Text-Independent Speaker Recognition: from Features to Supervectors

An Overview of Text-Independent Speaker Recognition: from Features to Supervectors An Overview of Text-Independent Speaker Recognition: from Features to Supervectors Tomi Kinnunen,a, Haizhou Li b a Department of Computer Science and Statistics, Speech and Image Processing Unit University

More information

TouchKit Software User manual for Windows 7 Version: 5.10.5

TouchKit Software User manual for Windows 7 Version: 5.10.5 TouchKit Software User manual for Windows 7 Version: 5.10.5 TouchKit V5.10.5 0 CONTENT CHAPTER 1. INSTALLING TOUCHKIT 2 CHAPTER 2. USING TOUCHKIT UTILITY...9 2.1 General...9 2.2 Tool...11 2.3 Setting...14

More information

Emotion Recognition Using Blue Eyes Technology

Emotion Recognition Using Blue Eyes Technology Emotion Recognition Using Blue Eyes Technology Prof. Sudan Pawar Shubham Vibhute Ashish Patil Vikram More Gaurav Sane Abstract We cannot measure the world of science in terms of progress and fact of development.

More information

Efficient Speaker Recognition for Mobile Devices

Efficient Speaker Recognition for Mobile Devices EVGENY KARPOV Efficient Speaker Recognition for Mobile Devices Publications of the University of Eastern Finland Dissertations in Forestry and Natural Sciences No 52 Academic Dissertation To be presented

More information

Innovative Tools and Technology to use during Aural Rehabilitation Therapy

Innovative Tools and Technology to use during Aural Rehabilitation Therapy Innovative Tools and Technology to use during Aural Rehabilitation Therapy Jodi Creighton, M.S.,CCC-A,LSLS Cert. AVT Cincinnati Children s Hospital Medical Center As one parent I know said, You are not

More information

Statistics in Face Recognition: Analyzing Probability Distributions of PCA, ICA and LDA Performance Results

Statistics in Face Recognition: Analyzing Probability Distributions of PCA, ICA and LDA Performance Results Statistics in Face Recognition: Analyzing Probability Distributions of PCA, ICA and LDA Performance Results Kresimir Delac 1, Mislav Grgic 2 and Sonja Grgic 2 1 Croatian Telecom, Savska 32, Zagreb, Croatia,

More information

Subjective SNR measure for quality assessment of. speech coders \A cross language study

Subjective SNR measure for quality assessment of. speech coders \A cross language study Subjective SNR measure for quality assessment of speech coders \A cross language study Mamoru Nakatsui and Hideki Noda Communications Research Laboratory, Ministry of Posts and Telecommunications, 4-2-1,

More information

Speech Recognition on Cell Broadband Engine UCRL-PRES-223890

Speech Recognition on Cell Broadband Engine UCRL-PRES-223890 Speech Recognition on Cell Broadband Engine UCRL-PRES-223890 Yang Liu, Holger Jones, John Johnson, Sheila Vaidya (Lawrence Livermore National Laboratory) Michael Perrone, Borivoj Tydlitat, Ashwini Nanda

More information

QAM Demodulation. Performance Conclusion. o o o o o. (Nyquist shaping, Clock & Carrier Recovery, AGC, Adaptive Equaliser) o o. Wireless Communications

QAM Demodulation. Performance Conclusion. o o o o o. (Nyquist shaping, Clock & Carrier Recovery, AGC, Adaptive Equaliser) o o. Wireless Communications 0 QAM Demodulation o o o o o Application area What is QAM? What are QAM Demodulation Functions? General block diagram of QAM demodulator Explanation of the main function (Nyquist shaping, Clock & Carrier

More information

Automatic Emotion Recognition from Speech

Automatic Emotion Recognition from Speech Automatic Emotion Recognition from Speech A PhD Research Proposal Yazid Attabi and Pierre Dumouchel École de technologie supérieure, Montréal, Canada Centre de recherche informatique de Montréal, Montréal,

More information

Online Diarization of Telephone Conversations

Online Diarization of Telephone Conversations Odyssey 2 The Speaker and Language Recognition Workshop 28 June July 2, Brno, Czech Republic Online Diarization of Telephone Conversations Oshry Ben-Harush, Itshak Lapidot, Hugo Guterman Department of

More information

At the beginning of my career as a desktop support manager, I searched everywhere

At the beginning of my career as a desktop support manager, I searched everywhere SEPTEMBER 2013 Desktop Support Metrics Written by Mike Hanson Data analysis by Jenny Rains At the beginning of my career as a desktop support manager, I searched everywhere for examples of industry-standard

More information

CIVIL Corpus: Voice Quality for Forensic Speaker Comparison

CIVIL Corpus: Voice Quality for Forensic Speaker Comparison CIVIL Corpus: Voice Quality for Forensic Speaker Comparison Eugenia San Segundo Helena Alves Marianela Fernández Trinidad Phonetics Lab. CSIC CILC2013 Alicante 15 Marzo 2013 CIVIL Project Cualidad Individual

More information

USER AUTHENTICATION USING ON-LINE SIGNATURE AND SPEECH

USER AUTHENTICATION USING ON-LINE SIGNATURE AND SPEECH USER AUTHENTICATION USING ON-LINE SIGNATURE AND SPEECH By Stephen Krawczyk A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE

More information

LEARNING FEATURE MAPPING USING DEEP NEURAL NETWORK BOTTLENECK FEATURES FOR DISTANT LARGE VOCABULARY SPEECH RECOGNITION

LEARNING FEATURE MAPPING USING DEEP NEURAL NETWORK BOTTLENECK FEATURES FOR DISTANT LARGE VOCABULARY SPEECH RECOGNITION LEARNING FEATURE MAPPING USING DEEP NEURAL NETWORK BOTTLENECK FEATURES FOR DISTANT LARGE VOCABULARY SPEECH RECOGNITION Ivan Himawan 1, Petr Motlicek 1, David Imseng 1, Blaise Potard 1, Namhoon Kim 2, Jaewon

More information

USB Smart Power Sensor

USB Smart Power Sensor 50Ω -30 dbm to +20 dbm, 9 khz to 4000 MHz The Big Deal Low cost USB HID device compatible with 32/64 Bit operating systems Includes Measurement Application GUI (Graphical User Interface) software with

More information

Disassembling a Windows Wave File (.wav)

Disassembling a Windows Wave File (.wav) Disassembling a Windows Wave File (.wav) The aim of this project was to create a nice sine wave as a test signal to trace the function of the electronic circuit of a tube amplifier by oscilloscope: By

More information

Various Technics of Liquids and Solids Level Measurements. (Part 3)

Various Technics of Liquids and Solids Level Measurements. (Part 3) (Part 3) In part one of this series of articles, level measurement using a floating system was discusses and the instruments were recommended for each application. In the second part of these articles,

More information

Online Filtering for Radar Detection of Meteors

Online Filtering for Radar Detection of Meteors 1, Gustavo O. Alves 1, José M. Seixas 1, Fernando Marroquim 2, Cristina S. Vianna 2, Helio Takai 3 1 Signal Processing Laboratory, COPPE/Poli, Federal University of Rio de Janeiro, Brazil. 2 Physics Institute,

More information

Capacity Limits of MIMO Channels

Capacity Limits of MIMO Channels Tutorial and 4G Systems Capacity Limits of MIMO Channels Markku Juntti Contents 1. Introduction. Review of information theory 3. Fixed MIMO channels 4. Fading MIMO channels 5. Summary and Conclusions References

More information

The Effect of Long-Term Use of Drugs on Speaker s Fundamental Frequency

The Effect of Long-Term Use of Drugs on Speaker s Fundamental Frequency The Effect of Long-Term Use of Drugs on Speaker s Fundamental Frequency Andrey Raev 1, Yuri Matveev 1, Tatiana Goloshchapova 2 1 Speech Technology Center, St. Petersburg, RUSSIA {raev, matveev}@speechpro.com

More information

Experiments with Signal-Driven Symbolic Prosody for Statistical Parametric Speech Synthesis

Experiments with Signal-Driven Symbolic Prosody for Statistical Parametric Speech Synthesis Experiments with Signal-Driven Symbolic Prosody for Statistical Parametric Speech Synthesis Fabio Tesser, Giacomo Sommavilla, Giulio Paci, Piero Cosi Institute of Cognitive Sciences and Technologies, National

More information

The Effect of Voice over IP Transmission Degradations on MAP-EM-GMM Speaker Verification Performance

The Effect of Voice over IP Transmission Degradations on MAP-EM-GMM Speaker Verification Performance ARCHIVES OF ACOUSTICS Vol. 40, No. 3, pp. 407 417 (2015) Copyright c 2015 by PAN IPPT DOI: 10.1515/aoa-2015-0042 The Effect of Voice over IP Transmission Degradations on MAP-EM-GMM Speaker Verification

More information

School Class Monitoring System Based on Audio Signal Processing

School Class Monitoring System Based on Audio Signal Processing C. R. Rashmi 1,,C.P.Shantala 2 andt.r.yashavanth 3 1 Department of CSE, PG Student, CIT, Gubbi, Tumkur, Karnataka, India. 2 Department of CSE, Vice Principal & HOD, CIT, Gubbi, Tumkur, Karnataka, India.

More information

Assessment of Camera Phone Distortion and Implications for Watermarking

Assessment of Camera Phone Distortion and Implications for Watermarking Assessment of Camera Phone Distortion and Implications for Watermarking Aparna Gurijala, Alastair Reed and Eric Evans Digimarc Corporation, 9405 SW Gemini Drive, Beaverton, OR 97008, USA 1. INTRODUCTION

More information

Does Affirmative Action Create Educational Mismatches in Law School?

Does Affirmative Action Create Educational Mismatches in Law School? Does Affirmative Action Create Educational Mismatches in Law School? DOUG WILLIAMS SEPTEMBER 21, 2012 Mismatch Hypothesis A Student Will Learn More if Her Credentials Are Similar to Those of Her Median

More information

Publication List. Chen Zehua Department of Statistics & Applied Probability National University of Singapore

Publication List. Chen Zehua Department of Statistics & Applied Probability National University of Singapore Publication List Chen Zehua Department of Statistics & Applied Probability National University of Singapore Publications Journal Papers 1. Y. He and Z. Chen (2014). A sequential procedure for feature selection

More information

Scandinavian Dialect Syntax Transnational collaboration, data collection, and resource development

Scandinavian Dialect Syntax Transnational collaboration, data collection, and resource development Scandinavian Dialect Syntax Transnational collaboration, data collection, and resource development Janne Bondi Johannessen, Signe Laake, Kristin Hagen, Øystein Alexander Vangsnes, Tor Anders Åfarli, Arne

More information

Non-parametric score normalization for biometric verification systems

Non-parametric score normalization for biometric verification systems Non-parametric score normalization for biometric verification systems Vitomir Štruc, Jerneja Žganec Gros 2, and Nikola Pavešić Faculty of Electrical Engineering, University of Ljubljana, Tržaška 25, Ljubljana,

More information

SOURCE SCANNER IDENTIFICATION FOR SCANNED DOCUMENTS. Nitin Khanna and Edward J. Delp

SOURCE SCANNER IDENTIFICATION FOR SCANNED DOCUMENTS. Nitin Khanna and Edward J. Delp SOURCE SCANNER IDENTIFICATION FOR SCANNED DOCUMENTS Nitin Khanna and Edward J. Delp Video and Image Processing Laboratory School of Electrical and Computer Engineering Purdue University West Lafayette,

More information

ANALYZER BASICS WHAT IS AN FFT SPECTRUM ANALYZER? 2-1

ANALYZER BASICS WHAT IS AN FFT SPECTRUM ANALYZER? 2-1 WHAT IS AN FFT SPECTRUM ANALYZER? ANALYZER BASICS The SR760 FFT Spectrum Analyzer takes a time varying input signal, like you would see on an oscilloscope trace, and computes its frequency spectrum. Fourier's

More information

Iteration 3 Kick Off, Domain Model Refinement. Curt Clifton Rose-Hulman Institute of Technology

Iteration 3 Kick Off, Domain Model Refinement. Curt Clifton Rose-Hulman Institute of Technology Iteration 3 Kick Off, Domain Model Refinement Curt Clifton Rose-Hulman Institute of Technology 2/3 Course Evaluation Results Lecture Pace 15 10 5 0 Much too slow Somewhat too slow Somewhat too fast Much

More information

CCNY. BME I5100: Biomedical Signal Processing. Linear Discrimination. Lucas C. Parra Biomedical Engineering Department City College of New York

CCNY. BME I5100: Biomedical Signal Processing. Linear Discrimination. Lucas C. Parra Biomedical Engineering Department City College of New York BME I5100: Biomedical Signal Processing Linear Discrimination Lucas C. Parra Biomedical Engineering Department CCNY 1 Schedule Week 1: Introduction Linear, stationary, normal - the stuff biology is not

More information

Formant Bandwidth and Resilience of Speech to Noise

Formant Bandwidth and Resilience of Speech to Noise Formant Bandwidth and Resilience of Speech to Noise Master Thesis Leny Vinceslas August 5, 211 Internship for the ATIAM Master s degree ENS - Laboratoire Psychologie de la Perception - Hearing Group Supervised

More information

Computer Networks and Internets, 5e Chapter 6 Information Sources and Signals. Introduction

Computer Networks and Internets, 5e Chapter 6 Information Sources and Signals. Introduction Computer Networks and Internets, 5e Chapter 6 Information Sources and Signals Modified from the lecture slides of Lami Kaya (LKaya@ieee.org) for use CECS 474, Fall 2008. 2009 Pearson Education Inc., Upper

More information

FACE authentication remains a challenging problem because

FACE authentication remains a challenging problem because IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY, VOL. X, NO. X, XXXXXXX 20XX 1 Cross-pollination of normalisation techniques from speaker to face authentication using Gaussian mixture models Roy

More information

TECHNOLOGY WHITEPAPER

TECHNOLOGY WHITEPAPER TECHNOLOGY WHITEPAPER ArmorVox10 Targets Call Center Fraud How to provide fast voice-file cross-matching to detect and track fraud in call centers. AURAYA SYSTEMS One Tara Boulevard Nashua, New Hampshire

More information

Probability and Random Variables. Generation of random variables (r.v.)

Probability and Random Variables. Generation of random variables (r.v.) Probability and Random Variables Method for generating random variables with a specified probability distribution function. Gaussian And Markov Processes Characterization of Stationary Random Process Linearly

More information

Section 6 Fire Detection and Alarm Systems Russell Porteous Chief Executive Officer Firewize Services

Section 6 Fire Detection and Alarm Systems Russell Porteous Chief Executive Officer Firewize Services Section 6 Fire Detection and Alarm Systems Russell Porteous Chief Executive Officer Firewize Services General Information Section 6 of AS1851-2012 covers: Fire Detection and Alarms Systems Electrical Detection

More information

Appendix A. CMS(Client Management Software)

Appendix A. CMS(Client Management Software) Appendix A. CMS(Client Management Software) A-1. Install CMS for Windows PC CMS is a program for communication between DVR and PC to control signal and video. Insert the enclosed CD, and go to CD-ROM Drive

More information

FACE RECOGNITION BASED ATTENDANCE MARKING SYSTEM

FACE RECOGNITION BASED ATTENDANCE MARKING SYSTEM Available Online at www.ijcsmc.com International Journal of Computer Science and Mobile Computing A Monthly Journal of Computer Science and Information Technology IJCSMC, Vol. 3, Issue. 2, February 2014,

More information

MIC - Detecting Novel Associations in Large Data Sets. by Nico Güttler, Andreas Ströhlein and Matt Huska

MIC - Detecting Novel Associations in Large Data Sets. by Nico Güttler, Andreas Ströhlein and Matt Huska MIC - Detecting Novel Associations in Large Data Sets by Nico Güttler, Andreas Ströhlein and Matt Huska Outline Motivation Method Results Criticism Conclusions Motivation - Goal Determine important undiscovered

More information

Clustering. 15-381 Artificial Intelligence Henry Lin. Organizing data into clusters such that there is

Clustering. 15-381 Artificial Intelligence Henry Lin. Organizing data into clusters such that there is Clustering 15-381 Artificial Intelligence Henry Lin Modified from excellent slides of Eamonn Keogh, Ziv Bar-Joseph, and Andrew Moore What is Clustering? Organizing data into clusters such that there is

More information

Selecting RJ Bandwidth in EZJIT Plus Software

Selecting RJ Bandwidth in EZJIT Plus Software Selecting RJ Bandwidth in EZJIT Plus Software Application Note 1577 Introduction Separating jitter into its random and deterministic components (called RJ/DJ separation ) is a relatively new technique

More information

Lecture 3: Linear methods for classification

Lecture 3: Linear methods for classification Lecture 3: Linear methods for classification Rafael A. Irizarry and Hector Corrada Bravo February, 2010 Today we describe four specific algorithms useful for classification problems: linear regression,

More information

Writer Identification for Smart Meeting Room Systems

Writer Identification for Smart Meeting Room Systems Writer Identification for Smart Meeting Room Systems Marcus Liwicki 1, Andreas Schlapbach 1, Horst Bunke 1, Samy Bengio 2, Johnny Mariéthoz 2, and Jonas Richiardi 3 1 Department of Computer Science, University

More information

FPGA Implementation of Human Behavior Analysis Using Facial Image

FPGA Implementation of Human Behavior Analysis Using Facial Image RESEARCH ARTICLE OPEN ACCESS FPGA Implementation of Human Behavior Analysis Using Facial Image A.J Ezhil, K. Adalarasu Department of Electronics & Communication Engineering PSNA College of Engineering

More information

FREE TV AUSTRALIA OPERATIONAL PRACTICE OP 60 Multi-Channel Sound Track Down-Mix and Up-Mix Draft Issue 1 April 2012 Page 1 of 6

FREE TV AUSTRALIA OPERATIONAL PRACTICE OP 60 Multi-Channel Sound Track Down-Mix and Up-Mix Draft Issue 1 April 2012 Page 1 of 6 Page 1 of 6 1. Scope. This operational practice sets out the requirements for downmixing 5.1 and 5.0 channel surround sound audio mixes to 2 channel stereo. This operational practice recommends a number

More information

Design of Experiments for Analytical Method Development and Validation

Design of Experiments for Analytical Method Development and Validation Design of Experiments for Analytical Method Development and Validation Thomas A. Little Ph.D. 2/12/2014 President Thomas A. Little Consulting 12401 N Wildflower Lane Highland, UT 84003 1-925-285-1847 drlittle@dr-tom.com

More information

Abstract. 1. Introduction. 1.1. Methodology

Abstract. 1. Introduction. 1.1. Methodology Fingerprint Recognition System Performance in the Maritime Environment Hourieh Fakourfar 1, Serge Belongie 2* 1 Department of Electrical and Computer Engineering, and 2 Department of Computer Science and

More information

KVM Cable Length Best Practices Guide

KVM Cable Length Best Practices Guide Infrastructure Management & Monitoring for Business-Critical Continuity TM KVM Cable Length Best Practices Guide What Customers Need to Know About Cable Length and Video Quality Cable Length and Video

More information

SPEAKER IDENTIFICATION & VERIFICATION

SPEAKER IDENTIFICATION & VERIFICATION SPEAKER IDENTIFICATION & VERIFICATION A Position Paper by: Bruce Balentine, Enterprise Integration Group bruce@eiglabs.com Introduction and Overview This document is a position paper prepared by Bruce

More information

USB Smart Power Sensor

USB Smart Power Sensor Low Power Measurement 50Ω -45 dbm to +10 dbm, 50 to 6000 MHz The Big Deal USB HID device compatible with 32/64 Bit operating systems Includes GUI with measurement applications software, simplifying complex

More information

Why Is This Topic So Important? Communication Styles: The Secret of Flexible Behavior. Considerations Regarding Communication

Why Is This Topic So Important? Communication Styles: The Secret of Flexible Behavior. Considerations Regarding Communication Styles: The Secret of Flexible Behavior Lisa O Connor, M.A. ASHA Certified Speech-Language Pathologist Why Is This Topic So Important? We spend a staggering amount of time communicating. We can all benefit

More information

Unlocking Value from. Patanjali V, Lead Data Scientist, Tiger Analytics Anand B, Director Analytics Consulting,Tiger Analytics

Unlocking Value from. Patanjali V, Lead Data Scientist, Tiger Analytics Anand B, Director Analytics Consulting,Tiger Analytics Unlocking Value from Patanjali V, Lead Data Scientist, Anand B, Director Analytics Consulting, EXECUTIVE SUMMARY Today a lot of unstructured data is being generated in the form of text, images, videos

More information

Some Thoughts on Climate Change and Software Engineering Research

Some Thoughts on Climate Change and Software Engineering Research Some Thoughts on Climate Change and Software Engineering Research Lin Liu 1, He Zhang 1, Sheikh Iqbal Ahamed 2 1 School of Software, Tsinghua University, Beijing, China 2 Department of Mathematics, Statistics

More information

Field Calibration Software

Field Calibration Software SIGNAL HOUND Field Calibration Software User s Manual Version 1.1.0 7/8/2016 This information is being released into the public domain in accordance with the Export Administration Regulations 15 CFR 734

More information

Unsupervised and supervised dimension reduction: Algorithms and connections

Unsupervised and supervised dimension reduction: Algorithms and connections Unsupervised and supervised dimension reduction: Algorithms and connections Jieping Ye Department of Computer Science and Engineering Evolutionary Functional Genomics Center The Biodesign Institute Arizona

More information

Implementing an In-Service, Non- Intrusive Measurement Device in Telecommunication Networks Using the TMS320C31

Implementing an In-Service, Non- Intrusive Measurement Device in Telecommunication Networks Using the TMS320C31 Disclaimer: This document was part of the First European DSP Education and Research Conference. It may have been written by someone whose native language is not English. TI assumes no liability for the

More information

US-Key New generation of High performances Ultrasonic device

US-Key New generation of High performances Ultrasonic device US-Key New generation of High performances Ultrasonic device US-Key connected to a laptop computer US-Key Ultrasound device single channel Features USB2 High Speed connection Ultralow noise preamplifier

More information