# Robust DNN-based VAD augmented with phone entropy based rejection of background speech

Size: px
Start display at page:

## Transcription

2 its output in the entropy calculation which is a crucial part of our proposed method. We conducted preliminary experiments and confirmed that there was little difference in the performance of these two approaches. We now describe in detail our VAD algorithm. Suppose x(t) is an acoustic feature vector at the t-th time frame and W l, b l are respectively the l-th layer s weight matrix and bias vector of an acoustic model DNN with L layers, then the posterior probability is calculated as follows: The 1-st hidden layer s output is calculated by h 1(t) = W 1x(t) + b 1, (1) o 1(t) = g 1 (h 1(t)), (2) and the l = {2,, L}-th layers output is calculated by h l (t) = W l o l 1 (t) + b l, (3) o l (t) = g l (h l (t)), (4) where g l ( ) is a non-linear activation function for the l-th layer. We used the sigmoid function for l = {1,, L 1}-th layers defined by g l (y) = exp( y), (5) and the identity function for the L-th layer. The final L-th layer s output is converted to posterior probabilities using the softmax function: p(i x(t)) = exp(oi L(t)), (6) i exp(o i L (t)) where o i L(t) represents the i-th component of vector o L (t). Then, the posterior probability of the speech hypothesis H 1 and non-speech hypothesis H 0 is calculated as follows: p(h 1 x(t)) = i S p(i x(t)), (7) p(h 0 x(t)) = i N p(i x(t)), (8) where S denotes the set of indices representing speech states and N represents the set of indices of silence states. If the following condition is met, we decide the t-th frame is a speech frame: p(h 1 x(t)) > p(h 0 x(t)). (9) In our method, the entropy based decision is also applied to speech frames classified by the above criterion. The entropy of each frame is calculated by e(t) = p(i x(t)) log p(i x(t)), (10) i S N so if the following condition is met, the t-th frame is identified as target speech and passed to the decoder: e(t) < τ. (11) A diagram of this algorithm is shown in Fig.1 As we mentioned in the introduction, the posterior probability of background speech could become close to a uniform distribution because of contamination by noise or reverberation so its entropy value becomes larger than clear utterances. We Figure 1: Diagram of proposed VAD method. show the waveform, manually labeled voice regions, posterior probability of speech and the entropy value of two utterances in Fig.2 and 3. Fig.2 is a plot of a clean and clear utterance. We can see that the entropy values do not become large. Fig.3 is an utterance corrupted by speech from the radio in a car environment. Both before and after the correct voice region, the posterior probability of speech becomes larger because of background speech. In that region, the entropy value becomes larger than those of the correct voice region. We plot the histograms of entropy of our development set in Fig.4 in order to see whether it is possible to classify background speech using the entropy value. Each frame of the development set is tagged as true positive, true negative, false alarm or false rejection by comparing labels generated by forced alignment. By manually checking several utterances from the development set, we confirm that most false alarms are caused by background speech. It is clear that the entropy value of frames tagged as false alarm are larger than other frames. We also plot the histogram of the moving average of entropy in Fig.5 and 6 because in [11] it is shown that averaging entropy over multiple frames makes it easier to discriminate a frame of speech or music. We expect that it works well in our background speech classification scenario too. However, averaging over multiple frames makes the histogram of false alarm frames close to true positive frames so we add the frame-wise entropybased decision criterion to reject background speech frames. If e(t) is greater than some threshold, the t-th frame is classified as background speech Experimental setup 3. Experiment We evaluated the conventional and proposed VAD method using the acoustic model DNN trained on 1200 hours of transcribed speech collected through our mobile voice search system. The conventional baseline method uses only Eq. (9) and the proposed method uses Eq. (9) and (11) for speech classification as shown in Fig.1. We select 20k utterances from map applications (Vehicle domain we defined in the introduction) which are different to those used in training. Then, we divide them equally into development and evaluation sets in such a way that each set does not contain utterances from the same period of time and from the same smartphone (each set has 10k utterances). In addition to these two sets, we prepared two reduced test sets (each a subset of the above 10k evaluation and development sets, respectively) to see the contribution of the VAD method to recog- 3664

3 Relative Frequency Waveform Labeled voice region Posterior probability of speech state of posterior (divided by 6.0) sec. Figure 2: Waveform, manually labeled voice region, posterior probability of speech state and entropy and of an utterance by a single speaker without background noise Figure 4: Histogram of entropy of development set. to a VAD process and classified frame-wise into speech or nonspeech. Then, the frame-wise VAD results are smoothed using a manually tuned finite-state automaton. After that, the segmented speech regions are passed to the decoder. Our decoder is an internally developed single-pass WFST decoder[12]. The language model is a tri-gram model trained using text queries of the Yahoo Japan search engine and transcriptions of mobile voice search queries. Other parameters are detailed in Table 1. Waveform Labeled voice region Posterior probability of speech state of posterior (divided by 6.0) sec. Figure 3: Waveform, manually labeled voice region, posterior probability of speech state and entropy of an utterance with background speech. nition accuracy. Speech recognition errors are caused by VAD errors or ASR decoder errors, and it is not trivial to separate these causes in general. In the reduced test sets, we chose utterances from the original test sets that are correctly recognized using manually labeled VAD boundaries. With these reduced test sets, we can estimate the contribution of our VAD method to recognition accuracy. We use two metrics to analyze performance. The first one is VAD frame error rate (FER) which is the number of frames misclassified divided by the total number of frames. The second one is phone Sentence Error Rate (SER). The reason for choosing SER is that our system is designed for mobile voice search in Japanese where the commonly used Word Error Rate (WER) metric does not always reflect the subjective performance by a user. This is because an error of one word may result in a completely different search result. The reason for using only phone information is because Japanese has 4 alphabets (kanji, hiragana, katakana and romaji) and one sentence can have multiple surface forms while having the same meaning therefore we normalized all surface forms to phones. The audio signal of each utterance in the test set is first sent Table 1: Parameters of the speech recognition system. name value Acoustic feature 40ch Filter Bank Splicing -5/+5 Number of units in hidden layers 1024 Number of hidden layers 5 Output state numbers 4003 Vocabulary size 1.3M 3.2. Results VAD FER of the development set is shown in Table 2. The best FER is observed when we set the entropy threshold to 7.0. At that operating point, the relative reduction in FER was 5.5%. VAD FER of the evaluation set is shown in Table 3 where the relative improvement was 2.4%. Table 2: VAD FER of the development set. Method threshold FER % baseline proposed The SER of the reduced test set is shown in Table 4. A relative reduction in SER of more than 10% was achieved. These results show that our proposed method can correctly recognize sentences that the baseline system could not. The SER of the whole test set is shown in Table 5. The reduction in SER on the development set was 4% and on the evaluation set was 2.2%. Note that the whole test set contains mis-recognized sentences 3665

5 [6] X.-L. Zhang and J. Wu, Deep belief networks based voice activity detection, IEEE Transactions on Audio, Speech, and Language Processing, vol. 21, no. 4, pp , April [7] Q. Wang, J. Du, X. Bao, Z. Wang, L. Dai, and C. Lee, A universal VAD based on jointly trained deep neural networks, in IN- TERSPEECH 2015, 16th Annual Conference of the International Speech Communication Association, Dresden, Germany, September 6-10, 2015, 2015, pp [8] I. Hwang, J. Sim, S. Kim, K. Song, and J. Chang, A statistical model-based voice activity detection using multiple dnns and noise awareness, in INTERSPEECH 2015, 16th Annual Conference of the International Speech Communication Association, Dresden, Germany, September 6-10, 2015, 2015, pp [9] J.-l. Shen, J.-w. Hung, and L.-s. Lee, Robust entropy-based endpoint detection for speech recognition in noisy environments. in ICSLP, vol. 98, 1998, pp [10] C. Yang and M. Hsieh, Robust endpoint detection for in-car speech recognition, in Sixth International Conference on Spoken Language Processing, ICSLP 2000, Beijing, China, October 16-20, 2000, 2000, pp [11] J. Ajmera, I. McCowan, and H. Bourlard, Speech/music segmentation using entropy and dynamism features in a HMM classification framework, Speech Communication, vol. 40, no. 3, pp , [12] K. Iso, E. Whittaker, T. Emori, and J. Miyake, Improvements in Japanese Voice Search, in INTERSPEECH 2012, 13th Annual Conference of the International Speech Communication Association, Portland, Oregon, USA, September 9-13, 2012, 2012, pp

### Robust Speech Recognition using DNN-HMM Acoustic Model Combining Noise-aware training with Spectral Subtraction

INTERSPEECH 2015 Robust Speech Recognition using DNN-HMM Acoustic Model Combining Noise-aware training with Spectral Subtraction Akihiro Abe, Kazumasa Yamamoto, Seiichi Nakagawa Department of Computer

### Calibration of Confidence Measures in Speech Recognition

Submitted to IEEE Trans on Audio, Speech, and Language, July 2010 1 Calibration of Confidence Measures in Speech Recognition Dong Yu, Senior Member, IEEE, Jinyu Li, Member, IEEE, Li Deng, Fellow, IEEE

### ADVANCES IN DEEP NEURAL NETWORK APPROACHES TO SPEAKER RECOGNITION

ADVANCES IN DEEP NEURAL NETWORK APPROACHES TO SPEAKER RECOGNITION Mitchell McLaren 1, Yun Lei 1, Luciana Ferrer 2 1 Speech Technology and Research Laboratory, SRI International, California, USA 2 Departamento

### PREDICTING SPEECH RECOGNITION CONFIDENCE USING DEEP LEARNING WITH WORD IDENTITY AND SCORE FEATURES

PREDICTING SPEECH RECOGNITION CONFIDENCE USING DEEP LEARNING WITH WORD IDENTITY AND SCORE FEATURES Po-Sen Huang, Kshitiz Kumar, Chaojun Liu, Yifan Gong, Li Deng Department of Electrical and Computer Engineering,

### A New Perspective on Combining GMM and DNN Frameworks for Speaker Adaptation

A New Perspective on Combining GMM and DNN Frameworks for Speaker Adaptation SLSP-2016 October 11-12 Natalia Tomashenko 1,2,3 natalia.tomashenko@univ-lemans.fr Yuri Khokhlov 3 khokhlov@speechpro.com Yannick

### Speech Recognition at ICSI: Broadcast News and beyond

Speech Recognition at ICSI: Broadcast News and beyond Dan Ellis International Computer Science Institute, Berkeley CA Outline 1 2 3 The DARPA Broadcast News task Aspects of ICSI

### Learning Methods in Multilingual Speech Recognition

Learning Methods in Multilingual Speech Recognition Hui Lin Department of Electrical Engineering University of Washington Seattle, WA 98125 linhui@u.washington.edu Li Deng, Jasha Droppo, Dong Yu, and Alex

### Semi-Supervised GMM and DNN Acoustic Model Training with Multi-system Combination and Confidence Re-calibration

INTERSPEECH 2013 Semi-Supervised GMM and DNN Acoustic Model Training with Multi-system Combination and Confidence Re-calibration Yan Huang, Dong Yu, Yifan Gong, and Chaojun Liu Microsoft Corporation, One

### Speech Emotion Recognition Using Support Vector Machine

Speech Emotion Recognition Using Support Vector Machine Yixiong Pan, Peipei Shen and Liping Shen Department of Computer Technology Shanghai JiaoTong University, Shanghai, China panyixiong@sjtu.edu.cn,

### Modeling function word errors in DNN-HMM based LVCSR systems

Modeling function word errors in DNN-HMM based LVCSR systems Melvin Jose Johnson Premkumar, Ankur Bapna and Sree Avinash Parchuri Department of Computer Science Department of Electrical Engineering Stanford

### Improvements to the Pruning Behavior of DNN Acoustic Models

Improvements to the Pruning Behavior of DNN Acoustic Models Matthias Paulik Apple Inc., Infinite Loop, Cupertino, CA 954 mpaulik@apple.com Abstract This paper examines two strategies that positively influence

### Modeling function word errors in DNN-HMM based LVCSR systems

Modeling function word errors in DNN-HMM based LVCSR systems Melvin Jose Johnson Premkumar, Ankur Bapna and Sree Avinash Parchuri Department of Computer Science Department of Electrical Engineering Stanford

### Autoregressive product of multi-frame predictions can improve the accuracy of hybrid models

Autoregressive product of multi-frame predictions can improve the accuracy of hybrid models Navdeep Jaitly 1, Vincent Vanhoucke 2, Geoffrey Hinton 1,2 1 University of Toronto 2 Google Inc. ndjaitly@cs.toronto.edu,

### WHEN THERE IS A mismatch between the acoustic

808 IEEE TRANSACTIONS ON AUDIO, SPEECH, AND LANGUAGE PROCESSING, VOL. 14, NO. 3, MAY 2006 Optimization of Temporal Filters for Constructing Robust Features in Speech Recognition Jeih-Weih Hung, Member,

### Lecture 1: Machine Learning Basics

1/69 Lecture 1: Machine Learning Basics Ali Harakeh University of Waterloo WAVE Lab ali.harakeh@uwaterloo.ca May 1, 2017 2/69 Overview 1 Learning Algorithms 2 Capacity, Overfitting, and Underfitting 3

### Lip reading: Japanese vowel recognition by tracking temporal changes of lip shape

Lip reading: Japanese vowel recognition by tracking temporal changes of lip shape Koshi Odagiri 1, and Yoichi Muraoka 1 1 Graduate School of Fundamental/Computer Science and Engineering, Waseda University,

### A study of speaker adaptation for DNN-based speech synthesis

A study of speaker adaptation for DNN-based speech synthesis Zhizheng Wu, Pawel Swietojanski, Christophe Veaux, Steve Renals, Simon King The Centre for Speech Technology Research (CSTR) University of Edinburgh,

### IEEE TRANSACTIONS ON AUDIO, SPEECH, AND LANGUAGE PROCESSING, VOL. 17, NO. 3, MARCH

IEEE TRANSACTIONS ON AUDIO, SPEECH, AND LANGUAGE PROCESSING, VOL. 17, NO. 3, MARCH 2009 423 Adaptive Multimodal Fusion by Uncertainty Compensation With Application to Audiovisual Speech Recognition George

### Word Segmentation of Off-line Handwritten Documents

Word Segmentation of Off-line Handwritten Documents Chen Huang and Sargur N. Srihari {chuang5, srihari}@cedar.buffalo.edu Center of Excellence for Document Analysis and Recognition (CEDAR), Department

### BUILDING CONTEXT-DEPENDENT DNN ACOUSTIC MODELS USING KULLBACK-LEIBLER DIVERGENCE-BASED STATE TYING

BUILDING CONTEXT-DEPENDENT DNN ACOUSTIC MODELS USING KULLBACK-LEIBLER DIVERGENCE-BASED STATE TYING Gábor Gosztolya 1, Tamás Grósz 1, László Tóth 1, David Imseng 2 1 MTA-SZTE Research Group on Artificial

### Unvoiced Landmark Detection for Segment-based Mandarin Continuous Speech Recognition

Unvoiced Landmark Detection for Segment-based Mandarin Continuous Speech Recognition Hua Zhang, Yun Tang, Wenju Liu and Bo Xu National Laboratory of Pattern Recognition Institute of Automation, Chinese

### Speech Segmentation Using Probabilistic Phonetic Feature Hierarchy and Support Vector Machines

Speech Segmentation Using Probabilistic Phonetic Feature Hierarchy and Support Vector Machines Amit Juneja and Carol Espy-Wilson Department of Electrical and Computer Engineering University of Maryland,

### A NOVEL SCHEME FOR SPEAKER RECOGNITION USING A PHONETICALLY-AWARE DEEP NEURAL NETWORK. Yun Lei Nicolas Scheffer Luciana Ferrer Mitchell McLaren

A NOVEL SCHEME FOR SPEAKER RECOGNITION USING A PHONETICALLY-AWARE DEEP NEURAL NETWORK Yun Lei Nicolas Scheffer Luciana Ferrer Mitchell McLaren Speech Technology and Research Laboratory, SRI International,

### Human Emotion Recognition From Speech

RESEARCH ARTICLE OPEN ACCESS Human Emotion Recognition From Speech Miss. Aparna P. Wanare*, Prof. Shankar N. Dandare *(Department of Electronics & Telecommunication Engineering, Sant Gadge Baba Amravati

### Segmental Conditional Random Fields with Deep Neural Networks as Acoustic Models for First-Pass Word Recognition

Segmental Conditional Random Fields with Deep Neural Networks as Acoustic Models for First-Pass Word Recognition Yanzhang He, Eric Fosler-Lussier Department of Computer Science and Engineering The hio

### Using dialogue context to improve parsing performance in dialogue systems

Using dialogue context to improve parsing performance in dialogue systems Ivan Meza-Ruiz and Oliver Lemon School of Informatics, Edinburgh University 2 Buccleuch Place, Edinburgh I.V.Meza-Ruiz@sms.ed.ac.uk,

### Phonetic- and Speaker-Discriminant Features for Speaker Recognition. Research Project

Phonetic- and Speaker-Discriminant Features for Speaker Recognition by Lara Stoll Research Project Submitted to the Department of Electrical Engineering and Computer Sciences, University of California

### Analysis of Emotion Recognition System through Speech Signal Using KNN & GMM Classifier

IOSR Journal of Electronics and Communication Engineering (IOSR-JECE) e-issn: 2278-2834,p- ISSN: 2278-8735.Volume 10, Issue 2, Ver.1 (Mar - Apr.2015), PP 55-61 www.iosrjournals.org Analysis of Emotion

### Likelihood-Maximizing Beamforming for Robust Hands-Free Speech Recognition

MITSUBISHI ELECTRIC RESEARCH LABORATORIES http://www.merl.com Likelihood-Maximizing Beamforming for Robust Hands-Free Speech Recognition Seltzer, M.L.; Raj, B.; Stern, R.M. TR2004-088 December 2004 Abstract

### INVESTIGATION OF UNSUPERVISED ADAPTATION OF DNN ACOUSTIC MODELS WITH FILTER BANK INPUT

INVESTIGATION OF UNSUPERVISED ADAPTATION OF DNN ACOUSTIC MODELS WITH FILTER BANK INPUT Takuya Yoshioka,, Anton Ragni, Mark J. F. Gales Cambridge University Engineering Department, Cambridge, UK NTT Communication

### DIRECT ADAPTATION OF HYBRID DNN/HMM MODEL FOR FAST SPEAKER ADAPTATION IN LVCSR BASED ON SPEAKER CODE

2014 IEEE International Conference on Acoustic, Speech and Signal Processing (ICASSP) DIRECT ADAPTATION OF HYBRID DNN/HMM MODEL FOR FAST SPEAKER ADAPTATION IN LVCSR BASED ON SPEAKER CODE Shaofei Xue 1

### Module 12. Machine Learning. Version 2 CSE IIT, Kharagpur

Module 12 Machine Learning 12.1 Instructional Objective The students should understand the concept of learning systems Students should learn about different aspects of a learning system Students should

### On the Formation of Phoneme Categories in DNN Acoustic Models

On the Formation of Phoneme Categories in DNN Acoustic Models Tasha Nagamine Department of Electrical Engineering, Columbia University T. Nagamine Motivation Large performance gap between humans and state-

### Assignment 1: Predicting Amazon Review Ratings

Assignment 1: Predicting Amazon Review Ratings 1 Dataset Analysis Richard Park r2park@acsmail.ucsd.edu February 23, 2015 The dataset selected for this assignment comes from the set of Amazon reviews for

### The NICT/ATR speech synthesis system for the Blizzard Challenge 2008

The NICT/ATR speech synthesis system for the Blizzard Challenge 2008 Ranniery Maia 1,2, Jinfu Ni 1,2, Shinsuke Sakai 1,2, Tomoki Toda 1,3, Keiichi Tokuda 1,4 Tohru Shimizu 1,2, Satoshi Nakamura 1,2 1 National

### QuickStroke: An Incremental On-line Chinese Handwriting Recognition System

QuickStroke: An Incremental On-line Chinese Handwriting Recognition System Nada P. Matić John C. Platt Λ Tony Wang y Synaptics, Inc. 2381 Bering Drive San Jose, CA 95131, USA Abstract This paper presents

### SEMI-SUPERVISED ENSEMBLE DNN ACOUSTIC MODEL TRAINING

SEMI-SUPERVISED ENSEMBLE DNN ACOUSTIC MODEL TRAINING Sheng Li 1, Xugang Lu 2, Shinsuke Sakai 1, Masato Mimura 1 and Tatsuya Kawahara 1 1 School of Informatics, Kyoto University, Sakyo-ku, Kyoto 606-8501,

### OCR for Arabic using SIFT Descriptors With Online Failure Prediction

OCR for Arabic using SIFT Descriptors With Online Failure Prediction Andrey Stolyarenko, Nachum Dershowitz The Blavatnik School of Computer Science Tel Aviv University Tel Aviv, Israel Email: stloyare@tau.ac.il,

### Analysis of Speech Recognition Models for Real Time Captioning and Post Lecture Transcription

Analysis of Speech Recognition Models for Real Time Captioning and Post Lecture Transcription Wilny Wilson.P M.Tech Computer Science Student Thejus Engineering College Thrissur, India. Sindhu.S Computer

### A Comparison of DHMM and DTW for Isolated Digits Recognition System of Arabic Language

A Comparison of DHMM and DTW for Isolated Digits Recognition System of Arabic Language Z.HACHKAR 1,3, A. FARCHI 2, B.MOUNIR 1, J. EL ABBADI 3 1 Ecole Supérieure de Technologie, Safi, Morocco. zhachkar2000@yahoo.fr.

### System Implementation for SemEval-2017 Task 4 Subtask A Based on Interpolated Deep Neural Networks

System Implementation for SemEval-2017 Task 4 Subtask A Based on Interpolated Deep Neural Networks 1 Tzu-Hsuan Yang, 2 Tzu-Hsuan Tseng, and 3 Chia-Ping Chen Department of Computer Science and Engineering

### Python Machine Learning

Python Machine Learning Unlock deeper insights into machine learning with this vital guide to cuttingedge predictive analytics Sebastian Raschka [ PUBLISHING 1 open source I community experience distilled

### Investigation on Mandarin Broadcast News Speech Recognition

Investigation on Mandarin Broadcast News Speech Recognition Mei-Yuh Hwang 1, Xin Lei 1, Wen Wang 2, Takahiro Shinozaki 1 1 Univ. of Washington, Dept. of Electrical Engineering, Seattle, WA 98195 USA 2

### Reducing Features to Improve Bug Prediction

Reducing Features to Improve Bug Prediction Shivkumar Shivaji, E. James Whitehead, Jr., Ram Akella University of California Santa Cruz {shiv,ejw,ram}@soe.ucsc.edu Sunghun Kim Hong Kong University of Science

### INPE São José dos Campos

INPE-5479 PRE/1778 MONLINEAR ASPECTS OF DATA INTEGRATION FOR LAND COVER CLASSIFICATION IN A NEDRAL NETWORK ENVIRONNENT Maria Suelena S. Barros Valter Rodrigues INPE São José dos Campos 1993 SECRETARIA

### AUTOMATIC DETECTION OF PROLONGED FRICATIVE PHONEMES WITH THE HIDDEN MARKOV MODELS APPROACH 1. INTRODUCTION

JOURNAL OF MEDICAL INFORMATICS & TECHNOLOGIES Vol. 11/2007, ISSN 1642-6037 Marek WIŚNIEWSKI *, Wiesława KUNISZYK-JÓŹKOWIAK *, Elżbieta SMOŁKA *, Waldemar SUSZYŃSKI * HMM, recognition, speech, disorders

### Role of Pausing in Text-to-Speech Synthesis for Simultaneous Interpretation

Role of Pausing in Text-to-Speech Synthesis for Simultaneous Interpretation Vivek Kumar Rangarajan Sridhar, John Chen, Srinivas Bangalore, Alistair Conkie AT&T abs - Research 180 Park Avenue, Florham Park,

### Australian Journal of Basic and Applied Sciences

AENSI Journals Australian Journal of Basic and Applied Sciences ISSN:1991-8178 Journal home page: www.ajbasweb.com Feature Selection Technique Using Principal Component Analysis For Improving Fuzzy C-Mean

### Experiments with SMS Translation and Stochastic Gradient Descent in Spanish Text Author Profiling

Experiments with SMS Translation and Stochastic Gradient Descent in Spanish Text Author Profiling Notebook for PAN at CLEF 2013 Andrés Alfonso Caurcel Díaz 1 and José María Gómez Hidalgo 2 1 Universidad

### arxiv: v1 [cs.cl] 2 Apr 2017

Word-Alignment-Based Segment-Level Machine Translation Evaluation using Word Embeddings Junki Matsuo and Mamoru Komachi Graduate School of System Design, Tokyo Metropolitan University, Japan matsuo-junki@ed.tmu.ac.jp,

### Noise-Adaptive Perceptual Weighting in the AMR-WB Encoder for Increased Speech Loudness in Adverse Far-End Noise Conditions

26 24th European Signal Processing Conference (EUSIPCO) Noise-Adaptive Perceptual Weighting in the AMR-WB Encoder for Increased Speech Loudness in Adverse Far-End Noise Conditions Emma Jokinen Department

### Rule Learning with Negation: Issues Regarding Effectiveness

Rule Learning with Negation: Issues Regarding Effectiveness Stephanie Chua, Frans Coenen, and Grant Malcolm University of Liverpool Department of Computer Science, Ashton Building, Ashton Street, L69 3BX

### Rule Learning With Negation: Issues Regarding Effectiveness

Rule Learning With Negation: Issues Regarding Effectiveness S. Chua, F. Coenen, G. Malcolm University of Liverpool Department of Computer Science, Ashton Building, Ashton Street, L69 3BX Liverpool, United

### Malicious User Suppression for Cooperative Spectrum Sensing in Cognitive Radio Networks using Dixon s Outlier Detection Method

Malicious User Suppression for Cooperative Spectrum Sensing in Cognitive Radio Networks using Dixon s Outlier Detection Method Sanket S. Kalamkar and Adrish Banerjee Department of Electrical Engineering

### Eli Yamamoto, Satoshi Nakamura, Kiyohiro Shikano. Graduate School of Information Science, Nara Institute of Science & Technology

ISCA Archive SUBJECTIVE EVALUATION FOR HMM-BASED SPEECH-TO-LIP MOVEMENT SYNTHESIS Eli Yamamoto, Satoshi Nakamura, Kiyohiro Shikano Graduate School of Information Science, Nara Institute of Science & Technology

### STUDIES WITH FABRICATED SWITCHBOARD DATA: EXPLORING SOURCES OF MODEL-DATA MISMATCH

STUDIES WITH FABRICATED SWITCHBOARD DATA: EXPLORING SOURCES OF MODEL-DATA MISMATCH Don McAllaster, Larry Gillick, Francesco Scattone, Mike Newman Dragon Systems, Inc. 320 Nevada Street Newton, MA 02160

### Mandarin Lexical Tone Recognition: The Gating Paradigm

Kansas Working Papers in Linguistics, Vol. 0 (008), p. 8 Abstract Mandarin Lexical Tone Recognition: The Gating Paradigm Yuwen Lai and Jie Zhang University of Kansas Research on spoken word recognition

### Course Outline. Course Grading. Where to go for help. Academic Integrity. EE-589 Introduction to Neural Networks NN 1 EE

EE-589 Introduction to Neural Assistant Prof. Dr. Turgay IBRIKCI Room # 305 (322) 338 6868 / 139 Wensdays 9:00-12:00 Course Outline The course is divided in two parts: theory and practice. 1. Theory covers

### Deep Neural Network Language Models

Deep Neural Network Language Models Ebru Arısoy, Tara N. Sainath, Brian Kingsbury, Bhuvana Ramabhadran IBM T.J. Watson Research Center Yorktown Heights, NY, 10598, USA {earisoy, tsainath, bedk, bhuvana}@us.ibm.com

### Corrective Feedback and Persistent Learning for Information Extraction

Corrective Feedback and Persistent Learning for Information Extraction Aron Culotta a, Trausti Kristjansson b, Andrew McCallum a, Paul Viola c a Dept. of Computer Science, University of Massachusetts,

### Class-Discriminative Weighted Distortion Measure for VQ-Based Speaker Identification

Class-Discriminative Weighted Distortion Measure for VQ-Based Speaker Identification Tomi Kinnunen and Ismo Kärkkäinen University of Joensuu, Department of Computer Science, P.O. Box 111, 80101 JOENSUU,

### Design Of An Automatic Speaker Recognition System Using MFCC, Vector Quantization And LBG Algorithm

Design Of An Automatic Speaker Recognition System Using MFCC, Vector Quantization And LBG Algorithm Prof. Ch.Srinivasa Kumar Prof. and Head of department. Electronics and communication Nalanda Institute

### CS 446: Machine Learning

CS 446: Machine Learning Introduction to LBJava: a Learning Based Programming Language Writing classifiers Christos Christodoulopoulos Parisa Kordjamshidi Motivation 2 Motivation You still have not learnt

### On the Combined Behavior of Autonomous Resource Management Agents

On the Combined Behavior of Autonomous Resource Management Agents Siri Fagernes 1 and Alva L. Couch 2 1 Faculty of Engineering Oslo University College Oslo, Norway siri.fagernes@iu.hio.no 2 Computer Science

### Unsupervised Learning of Word Semantic Embedding using the Deep Structured Semantic Model

Unsupervised Learning of Word Semantic Embedding using the Deep Structured Semantic Model Xinying Song, Xiaodong He, Jianfeng Gao, Li Deng Microsoft Research, One Microsoft Way, Redmond, WA 98052, U.S.A.

### Segregation of Unvoiced Speech from Nonspeech Interference

Technical Report OSU-CISRC-8/7-TR63 Department of Computer Science and Engineering The Ohio State University Columbus, OH 4321-1277 FTP site: ftp.cse.ohio-state.edu Login: anonymous Directory: pub/tech-report/27

### Evolutive Neural Net Fuzzy Filtering: Basic Description

Journal of Intelligent Learning Systems and Applications, 2010, 2: 12-18 doi:10.4236/jilsa.2010.21002 Published Online February 2010 (http://www.scirp.org/journal/jilsa) Evolutive Neural Net Fuzzy Filtering:

### Speaker recognition using universal background model on YOHO database

Aalborg University Master Thesis project Speaker recognition using universal background model on YOHO database Author: Alexandre Majetniak Supervisor: Zheng-Hua Tan May 31, 2011 The Faculties of Engineering,

### Switchboard Language Model Improvement with Conversational Data from Gigaword

Katholieke Universiteit Leuven Faculty of Engineering Master in Artificial Intelligence (MAI) Speech and Language Technology (SLT) Switchboard Language Model Improvement with Conversational Data from Gigaword

### Twitter Sentiment Classification on Sanders Data using Hybrid Approach

IOSR Journal of Computer Engineering (IOSR-JCE) e-issn: 2278-0661,p-ISSN: 2278-8727, Volume 17, Issue 4, Ver. I (July Aug. 2015), PP 118-123 www.iosrjournals.org Twitter Sentiment Classification on Sanders

### Multi-Lingual Text Leveling

Multi-Lingual Text Leveling Salim Roukos, Jerome Quin, and Todd Ward IBM T. J. Watson Research Center, Yorktown Heights, NY 10598 {roukos,jlquinn,tward}@us.ibm.com Abstract. Determining the language proficiency

### MULTILINGUAL INFORMATION ACCESS IN DIGITAL LIBRARY

MULTILINGUAL INFORMATION ACCESS IN DIGITAL LIBRARY Chen, Hsin-Hsi Department of Computer Science and Information Engineering National Taiwan University Taipei, Taiwan E-mail: hh_chen@csie.ntu.edu.tw Abstract

### Semi-Supervised Face Detection

Semi-Supervised Face Detection Nicu Sebe, Ira Cohen 2, Thomas S. Huang 3, Theo Gevers Faculty of Science, University of Amsterdam, The Netherlands 2 HP Research Labs, USA 3 Beckman Institute, University

### BAUM-WELCH TRAINING FOR SEGMENT-BASED SPEECH RECOGNITION. Han Shu, I. Lee Hetherington, and James Glass

BAUM-WELCH TRAINING FOR SEGMENT-BASED SPEECH RECOGNITION Han Shu, I. Lee Hetherington, and James Glass Computer Science and Artificial Intelligence Laboratory Massachusetts Institute of Technology Cambridge,

### CLASSIFICATION OF TEXT DOCUMENTS USING INTEGER REPRESENTATION AND REGRESSION: AN INTEGRATED APPROACH

ISSN: 0976-3104 Danti and Bhushan. ARTICLE OPEN ACCESS CLASSIFICATION OF TEXT DOCUMENTS USING INTEGER REPRESENTATION AND REGRESSION: AN INTEGRATED APPROACH Ajit Danti 1 and SN Bharath Bhushan 2* 1 Department

### Introduction to Ensemble Learning Featuring Successes in the Netflix Prize Competition

Introduction to Ensemble Learning Featuring Successes in the Netflix Prize Competition Todd Holloway Two Lecture Series for B551 November 20 & 27, 2007 Indiana University Outline Introduction Bias and

### DOMAIN MISMATCH COMPENSATION FOR SPEAKER RECOGNITION USING A LIBRARY OF WHITENERS. Elliot Singer and Douglas Reynolds

DOMAIN MISMATCH COMPENSATION FOR SPEAKER RECOGNITION USING A LIBRARY OF WHITENERS Elliot Singer and Douglas Reynolds Massachusetts Institute of Technology Lincoln Laboratory {es,dar}@ll.mit.edu ABSTRACT

### Body-Conducted Speech Recognition and its Application to Speech Support System

Body-Conducted Speech Recognition and its Application to Speech Support System 4 Shunsuke Ishimitsu Hiroshima City University Japan 1. Introduction In recent years, speech recognition systems have been

### Speech Recognition using Acoustic Landmarks and Binary Phonetic Feature Classifiers

Speech Recognition using Acoustic Landmarks and Binary Phonetic Feature Classifiers October 31, 2003 Amit Juneja Department of Electrical and Computer Engineering University of Maryland, College Park,

### A Neural Network GUI Tested on Text-To-Phoneme Mapping

A Neural Network GUI Tested on Text-To-Phoneme Mapping MAARTEN TROMPPER Universiteit Utrecht m.f.a.trompper@students.uu.nl Abstract Text-to-phoneme (T2P) mapping is a necessary step in any speech synthesis

### Learning Structural Correspondences Across Different Linguistic Domains with Synchronous Neural Language Models

Learning Structural Correspondences Across Different Linguistic Domains with Synchronous Neural Language Models Stephan Gouws and GJ van Rooyen MIH Medialab, Stellenbosch University SOUTH AFRICA {stephan,gvrooyen}@ml.sun.ac.za

### Probabilistic Latent Semantic Analysis

Probabilistic Latent Semantic Analysis Thomas Hofmann Presentation by Ioannis Pavlopoulos & Andreas Damianou for the course of Data Mining & Exploration 1 Outline Latent Semantic Analysis o Need o Overview

### Unsupervised Acoustic Model Training for Simultaneous Lecture Translation in Incremental and Batch Mode

Unsupervised Acoustic Model Training for Simultaneous Lecture Translation in Incremental and Batch Mode Diploma Thesis of Michael Heck At the Department of Informatics Karlsruhe Institute of Technology

### Miscommunication and error handling

CHAPTER 3 Miscommunication and error handling In the previous chapter, conversation and spoken dialogue systems were described from a very general perspective. In this description, a fundamental issue

### A Case Study: News Classification Based on Term Frequency

A Case Study: News Classification Based on Term Frequency Petr Kroha Faculty of Computer Science University of Technology 09107 Chemnitz Germany kroha@informatik.tu-chemnitz.de Ricardo Baeza-Yates Center

### Generative models and adversarial training

Day 4 Lecture 1 Generative models and adversarial training Kevin McGuinness kevin.mcguinness@dcu.ie Research Fellow Insight Centre for Data Analytics Dublin City University What is a generative model?

### Affective Classification of Generic Audio Clips using Regression Models

Affective Classification of Generic Audio Clips using Regression Models Nikolaos Malandrakis 1, Shiva Sundaram, Alexandros Potamianos 3 1 Signal Analysis and Interpretation Laboratory (SAIL), USC, Los

### TRANSFER LEARNING OF WEAKLY LABELLED AUDIO. Aleksandr Diment, Tuomas Virtanen

TRANSFER LEARNING OF WEAKLY LABELLED AUDIO Aleksandr Diment, Tuomas Virtanen Tampere University of Technology Laboratory of Signal Processing Korkeakoulunkatu 1, 33720, Tampere, Finland firstname.lastname@tut.fi

### CS Machine Learning

CS 478 - Machine Learning Projects Data Representation Basic testing and evaluation schemes CS 478 Data and Testing 1 Programming Issues l Program in any platform you want l Realize that you will be doing

### DNN ACOUSTIC MODELING WITH MODULAR MULTI-LINGUAL FEATURE EXTRACTION NETWORKS

DNN ACOUSTIC MODELING WITH MODULAR MULTI-LINGUAL FEATURE EXTRACTION NETWORKS Jonas Gehring 1 Quoc Bao Nguyen 1 Florian Metze 2 Alex Waibel 1,2 1 Interactive Systems Lab, Karlsruhe Institute of Technology;

### Softprop: Softmax Neural Network Backpropagation Learning

Softprop: Softmax Neural Networ Bacpropagation Learning Michael Rimer Computer Science Department Brigham Young University Provo, UT 84602, USA E-mail: mrimer@axon.cs.byu.edu Tony Martinez Computer Science

### Distributed Learning of Multilingual DNN Feature Extractors using GPUs

Distributed Learning of Multilingual DNN Feature Extractors using GPUs Yajie Miao, Hao Zhang, Florian Metze Language Technologies Institute, School of Computer Science, Carnegie Mellon University Pittsburgh,

### Product Feature-based Ratings foropinionsummarization of E-Commerce Feedback Comments

Product Feature-based Ratings foropinionsummarization of E-Commerce Feedback Comments Vijayshri Ramkrishna Ingale PG Student, Department of Computer Engineering JSPM s Imperial College of Engineering &

### ACOUSTIC EVENT DETECTION IN REAL LIFE RECORDINGS

ACOUSTIC EVENT DETECTION IN REAL LIFE RECORDINGS Annamaria Mesaros 1, Toni Heittola 1, Antti Eronen 2, Tuomas Virtanen 1 1 Department of Signal Processing Tampere University of Technology Korkeakoulunkatu

### arxiv: v2 [cs.cv] 30 Mar 2017

Domain Adaptation for Visual Applications: A Comprehensive Survey Gabriela Csurka arxiv:1702.05374v2 [cs.cv] 30 Mar 2017 Abstract The aim of this paper 1 is to give an overview of domain adaptation and

### Support Vector Machines for Speaker and Language Recognition

Support Vector Machines for Speaker and Language Recognition W. M. Campbell, J. P. Campbell, D. A. Reynolds, E. Singer, P. A. Torres-Carrasquillo MIT Lincoln Laboratory, 244 Wood Street, Lexington, MA

### The Good Judgment Project: A large scale test of different methods of combining expert predictions

The Good Judgment Project: A large scale test of different methods of combining expert predictions Lyle Ungar, Barb Mellors, Jon Baron, Phil Tetlock, Jaime Ramos, Sam Swift The University of Pennsylvania

### CSL465/603 - Machine Learning

CSL465/603 - Machine Learning Fall 2016 Narayanan C Krishnan ckn@iitrpr.ac.in Introduction CSL465/603 - Machine Learning 1 Administrative Trivia Course Structure 3-0-2 Lecture Timings Monday 9.55-10.45am