Takeda / Erdogan / Abut In-Vehicle Corpus and Signal Processing for Driver Behavior

1. Auflage 2009
ISBN: 978-0-387-79582-9
Verlag: Springer-Verlag
Format: PDF
Kopierschutz: Adobe DRM (»Systemvoraussetzungen)

E-Book, Englisch, 286 Seiten

In-Vehicle Corpus and Signal Processing for Driver Behavior is comprised of expanded papers from the third biennial DSPinCARS held in Istanbul in June 2007. The goal is to bring together scholars working on the latest techniques, standards, and emerging deployment on this central field of living at the age of wireless communications, smart vehicles, and human-machine-assisted safer and comfortable driving. Topics covered in this book include: improved vehicle safety; safe driver assistance systems; smart vehicles; wireless LAN-based vehicular location information processing; EEG emotion recognition systems; and new methods for predicting driving actions using driving signals. In-Vehicle Corpus and Signal Processing for Driver Behavior is appropriate for researchers, engineers, and professionals working in signal processing technologies, next generation vehicle design, and networks for mobile platforms.

Autoren/Hrsg.

Takeda, Kazuya

Erdogan, Hakan

Abut, Huseyin

Hansen, John

Inhaltsverzeichnis

1;Contents;6
2;Contributing Authors;9
3;Introduction;12
4;Improved Vehicle Safety and How Technology Will Get Us There, Hopefully;15
4.1;1.1 Introduction;15
4.2;1.2 Highway Safety;16
4.3;1.3 Drivers;16
4.4;1.4 Attention, Perception, and System Interfaces;17
4.5;1.5 In-Vehicle System Technologies;19
4.6;1.6 Concerns for Driver Distraction;19
4.7;1.7 Conclusions;20
4.8;References;21
5;New Concepts on Safe Driver-Assistance Systems;23
5.1;2.1 Introduction;23
5.2;2.2 Driver-Assistance Systems in the Market;24
5.3;2.3 Driver-Adaptive Assistance Systems;27
5.4;2.4 Driver Assistance with Cooperation Among Vehciles;30
5.5;2.5 Discussion;35
5.6;2.6 Conclusion;35
5.7;References;36
6;Real-World Data Collection with ‘‘UYANIK’’;37
6.1;3.1 Introduction;37
6.2;3.2 ‘‘Uyanik’’ and Sensors;39
6.3;3.3 Tasks and Pains;42
6.4;3.4 Signal Samples;44
6.5;3.5 Gains are Coming: Part 1;46
6.5.1;3.5.1 Audio-Visual Speech Recognition in Vehicular Noise Using a Multi-classifier Approach by H. Karabalkan and H. Erdogbrevean;47
6.5.2;3.5.2 Graphical Model-Based Facial Feature Point Tracking in a Vehicle Environment by S. Coscedilar;50
6.5.3;3.5.3 3D Head Tracking Using Normal Flow Constraints in a Vehicle Environment by B. Akan;51
6.6;3.6 Gains are Coming: Part 2;54
6.6.1;3.6.1 Pedal Engagement Behavior of Drivers by M. Karaca, M. Abbak, and M.G. Uzunbascedil;54
6.6.2;3.6.2 Speaker Verification and Fingerprint Recognition by K. Eritmen, M. Imamogbrevelu, and Ç. Karabat;55
6.7;3.7 Conclusions and Future Work;56
6.8;References;57
7;On-Going Data Collection of Driving Behavior Signals;58
7.1;4.1 Introduction;58
7.2;4.2 Design of Data Collection Vehicle;59
7.2.1;4.2.1 Vehicle;59
7.2.2;4.2.2 Microphones;60
7.2.3;4.2.3 Video Cameras;62
7.2.4;4.2.4 Sensors for Driving Operation Signals;62
7.2.5;4.2.5 Vehicle Status Sensors;62
7.2.6;4.2.6 Vehicle Position Sensors;63
7.2.7;4.2.7 Physiological Sensors;63
7.2.8;4.2.8 Synchronous Recording System;63
7.3;4.3 Data Collection;64
7.3.1;4.3.1 Tasks;65
7.3.2;4.3.2 Examples of Driving Data;66
7.4;4.4 Conclusion and Future Work;66
7.5;References;66
8;UTDrive: The Smart Vehicle Project;68
8.1;5.1 Introduction;68
8.2;5.2 Multi-Modal Data Acquisition;70
8.2.1;5.2.1 Audio;70
8.2.2;5.2.2 Video;71
8.2.3;5.2.3 CAN-Bus Information;72
8.2.4;5.2.4 Transducers and Extensive Components;72
8.2.5;5.2.5 Data Acquisition Unit (DAC);73
8.3;5.3 Data Collection Protocol;74
8.4;5.4 Driving Signals;75
8.5;5.5 Driver Distractions;76
8.6;5.6 Driver Behavior Modeling;78
8.7;5.7 Transcription Convention;78
8.8;5.8 Conclusion and Future Work;79
8.9;References;79
9;Wireless Lan-Based Vehicular Location Information Processing;81
9.1;6.1 Introduction;81
9.2;6.2 Localization;82
9.2.1;6.2.1 Wireless LAN-Based Localization;82
9.2.2;6.2.2 Proximity Approach;83
9.2.3;6.2.3 Triangulation Approach;83
9.2.4;6.2.4 Scene Analysis Approach;84
9.2.5;6.2.5 Accuracy in Outdoor Configuration;84
9.3;6.3 Orientation Estimation;85
9.3.1;6.3.1 Difference of Signal Strength;85
9.3.2;6.3.2 Orientation Estimation in Vehicle;86
9.4;6.4 Metropolitan-Scale Localization;88
9.4.1;6.4.1 Feasibility of Metropolitan-Scale Localization;88
9.4.2;6.4.2 Locky.jp;90
9.4.3;6.4.3 Current status;91
9.4.4;6.4.4 Related Projects;91
9.4.5;6.4.5 On-Going and Future Work;92
9.5;6.5 Conclusion;93
9.6;References;93
10;Perceptually Optimized Packet Scheduling for Robust Real-Time Intervehicle Video Communications;95
10.1;7.1 Introduction;95
10.2;7.2 The Inter-Vehicle Video Communications Scenario;97
10.2.1;7.2.1 The H.264 Video Coding Standard;97
10.2.2;7.2.2 Analysis-by-Synthesis Distortion Estimation for Video Packets;97
10.2.3;7.2.3 Multimedia Communications over 802.11;99
10.3;7.3 The Perceptually Optimized Packet Scheduling Algorithm;100
10.4;7.4 Experimental Setup;101
10.5;7.5 Results;102
10.6;7.6 Conclusions;107
10.7;References;107
11;Machine Learning Systems for Detecting Driver Drowsiness;109
11.1;8.1 Introduction;109
11.2;8.2 Methods;111
11.2.1;8.2.1 Driving Task;111
11.2.2;8.2.2 Head Movement Measures;112
11.2.3;8.2.3 Facial Action Classifiers;112
11.3;8.3 Results;114
11.3.1;8.3.1 Facial Action Signals;115
11.3.2;8.3.2 Drowsiness Prediction;116
11.3.3;8.3.3 Coupling of Behaviors;119
11.4;8.4 Conclusions and Future Work;120
11.5;References;122
12;Extraction of Pedestrian Regions Using Histogram and Locally Estimated Feature Distribution;123
12.1;9.1 Introduction;123
12.2;9.2 Related Research;125
12.3;9.3 Kernel Density Estimator With Bayesian Discriminant Function;126
12.3.1;9.3.1 Region of Interest for Processing;126
12.3.2;9.3.2 Modifying Probability Distribution Function;126
12.3.3;9.3.3 Category Estimation by Histograms;128
12.3.4;9.3.4 Kernel Density Estimation with Proposed Preprocessing;128
12.4;9.4 Pre-Experiment With Gaussian Shape Distribution;130
12.5;9.5 Shape-Dependent Probability Map Template;130
12.5.1;9.5.1 Experiment of Criterion Performance;133
12.6;9.6 Conclusion;135
12.7;References;135
13;EEG Emotion Recognition System;137
13.1;10.1 Introduction;137
13.2;10.2 Emotional Data Collection;138
13.2.1;10.2.1 Experimental Setup;138
13.2.2;10.2.2 Psychological Experiments;139
13.3;10.3 Feature Extraction;140
13.3.1;10.3.1 Selection of Experiment Parameters;141
13.3.2;10.3.2 RVM Model;141
13.4;10.4 Experimental Results;142
13.5;10.5 Conclusions;146
13.6;References;147
14;Three-Dimensional Ultrasound Imaging in Air for Parking and Pedestrian Protection;148
14.1;11.1 Motivation - Why Ultrasound?;148
14.2;11.2 Signal Model;149
14.3;11.3 Image Generation;151
14.3.1;11.3.1 System Setup;152
14.3.2;11.3.2 Data Processing;152
14.3.3;11.3.3 Beamforming;153
14.3.4;11.3.4 Evaluation Criteria;153
14.4;11.4 Experiments and Discussion of Results;154
14.4.1;11.4.1 Rough Surface Structure-Continuous Response;154
14.4.2;11.4.2 Smooth Surface - Specular Response;156
14.5;11.5 Conclusions;157
14.6;References;158
15;A New Method for Evaluating Mental Work Load In n-Back Tasks;159
15.1;12.1 Introduction;160
15.2;12.2 Model of Eye Movement;161
15.2.1;12.2.1 Model of VOR;161
15.2.2;12.2.2 Model Identification;162
15.2.3;12.2.3 Identification Results;163
15.3;12.3 Method Of Experiment;164
15.3.1;12.3.1 Experiment Procedure;164
15.3.2;12.3.2 Results of Experiment;166
15.4;12.4 Conclusion;168
15.5;References;168
16;Estimation of Acoustic Microphone Vocal Tract Parameters from Throat Microphone Recordings;170
16.1;13.1 Introduction;170
16.2;13.2 Acoustic-Throat Correlation Model;172
16.2.1;13.2.1 Vector Quantization-Based Estimator;173
16.2.2;13.2.2 Hidden Markov Model-Based Estimator;174
16.3;13.3 Experimental Results;175
16.4;13.4 Conclusions;177
16.5;References;177
17;Cross-Probability Model Based on Gmm for Feature Vector Normalization;179
17.1;14.1 Introduction;179
17.2;14.2 Memlin Overview;181
17.2.1;14.2.1 MEMLIN Approximations;181
17.2.2;14.2.2 MEMLIN Enhancement;183
17.3;14.3 Cross-Probability Model Performance;183
17.4;14.4 Cross-Probability Model Based on GMM;185
17.4.1;14.4.1 The E Step;186
17.4.2;14.4.2 The M Step;187
17.5;14.5 Normalized Space Acoustic Models;188
17.6;14.6 Discussion of Results;188
17.6.1;14.6.1 Results with SpeechDat Car Database;188
17.6.2;14.6.2 Results with Aurora2 Database;191
17.7;14.7 Conclusions;192
17.8;References;192
18;Robust Feature Combination for Speech Recognition Using Linear Microphone Array in a Car;194
18.1;15.1 Introduction;194
18.2;15.2 MFCC Average and Variance Re-scaling;195
18.3;15.3 GMM-Based Variance Normalization;197
18.4;15.4 Hypothesis-Based Feature Combination of Multiple Inputs;197
18.5;15.5 Experimental Results;199
18.5.1;15.5.1 Database and Setup;199
18.5.2;15.5.2 MFCC Average;199
18.5.3;15.5.3 Hypothesis-Based Feature Combination;201
18.6;15.6 Conclusions;202
18.7;References;203
19;Prediction of Driving Actions from Driving Signals;204
19.1;16.1 Introduction;204
19.2;16.2 Driving Signals;205
19.2.1;16.2.1 Types of Driving Signals;205
19.2.2;16.2.2 Database;206
19.3;16.3 Predicting Driving Actions;206
19.3.1;16.3.1 Kinds of Driving Actions;206
19.3.2;16.3.2 Methodology of Driving Action Prediction;207
19.4;16.4 Experimental Data;208
19.4.1;16.4.1 Driving Action Labels;208
19.4.2;16.4.2 Training Data and Test Data;209
19.5;16.5 Results of Prediction Experiments;210
19.5.1;16.5.1 Prediction Performance for Different Driving Signal Input Durations (Experiment 1);210
19.5.2;16.5.2 Prediction Performance for Different Amounts of Training Data (Experiment 2);211
19.5.3;16.5.3 Prediction Performance that Considers Individuality of Driving;212
19.5.4;16.5.4 Prediction Performance Using Only One Signal out of five Driving Signals (Experiment 4);213
19.5.5;16.5.5 Prediction Performance Using Three Useful Signals out of Five;215
19.5.6;16.5.6 Prediction Performance Using Detailed Classification of Driving Actions (Experiment 6);215
19.6;16.6 Conclusion;217
19.7;References;217
20;Design of Audio-Visual Interface for Aiding Driver’s Voice Commands in Automotive Environment;218
20.1;17.1 Introduction;218
20.2;17.2 Visual Feature Extraction;219
20.3;17.3 SNR-Dependent Audio-Visual Information Combination;222
20.3.1;17.3.1 Acoustic Feature Extraction and Estimation of SNR;222
20.3.2;17.3.2 Audio-Visual Model Combination;223
20.4;17.4 Experiments and Results;224
20.5;17.5 Conclusion;225
20.6;References;225
21;Estimation of High-Variance Vehicular Noise;227
21.1;18.1 Introduction;227
21.2;18.2 Background;229
21.2.1;18.2.1 Statistical Noise Model;229
21.2.2;18.2.2 MMSE A Priori Noise Estimation;230
21.2.3;18.2.3 Noise Estimation with Speech Presence Uncertainty;230
21.2.4;18.2.4 MMSE A Posteriori Noise Estimation;231
21.2.5;18.2.5 Autoregressive Noise Adaptation;232
21.3;18.3 Estimation of High-Variance Noise;232
21.3.1;18.3.1 Speech Presence Probability with Low SNR;232
21.3.2;18.3.2 Proposed Noise Estimation Method;234
21.4;18.4 Experiments;234
21.5;18.5 Conclusion;237
21.6;References;237
22;Feature Compensation Employing Model Combination for Robust In-Vehicle Speech Recognition;239
22.1;19.1 Introduction;240
22.2;19.2 CU-Move Corpus;241
22.3;19.3 PCGMM-Based Feature Compensation;241
22.4;19.4 PCGMM-Based Method Employing Multiple Environmental Models;243
22.5;19.5 Noise Transition Model;244
22.6;19.6 Experimental Results;245
22.7;19.7 Conclusion;248
22.8;References;249
23;Index;250