Autonomous face recognition [Elektronische Ressource] / von Mou, Dengpan

Sujets

• Informatik

Informations

• Publié par: universitat_ulm
• Publié le: 01 janvier 2005
• Nombre de lectures: 17
• Langue: English
• Poids de l'ouvrage: 4 Mo

Extrait

Autonomous Face Recognition




Dissertation

zur Erlangung des akademischen Grades eines
Doktor-Ingenieurs (Dr.-Ing.)
der Fakultät für Ingenieurwissenschaften
der Universität Ulm


von
Mou, Dengpan
aus China


1. Gutachter: Prof. Dr.-Ing. Albrecht Rothermel
2. Gutachter: Prof. Dr. Heiko Neumann
Amtierender Dekan: Prof. Dr.-Ing. Hans-Jörg Pfleiderer
Datum der Promotion: 22. August, 2005

2005

II Abstract
As a booming technology, face recognition has been studied for many years and is expected
to be widely used in daily identification systems, communication systems, public security
systems, and in law enforcement systems.
Most state-of-the-art machine learning systems are based on the supervised learning theory
and image processing techniques, which require separate pre-training procedure for enrolling
every new face and updating existing faces. Therefore, an additional human supervisor is
normally required. Users’ cooperation is expected as well during the training phase.
However, a human vision system is far more intelligent. It has no difficulty to automatically
memorize the faces they have interacted with for future recognition. All the enrollments,
updates, and comparisons have been done completely in the brain without any outside
assistance. Although the biological reasons behind it are not clear until now, it is not hard to
imagine that the brains can combine all information that is useful for recognition, including
“image processing”, video context, logic deduction, experiences etc.
Inspired from the human vision system, we combined the conventional learning algorithms
and image processing algorithms with predefined rules to increase the intelligence of
machine recognition systems. As the first step, face detection is implemented by an
industrial image-based face detector combined with novel temporal differencing algorithms.
The face detection result, an industrial image-based classifier, temporal filtering and video
context related rules are all combined for face recognition. The database can be constructed
online and be adapted automatically according to the update rules. State machine is
introduced to keep the system running automatically and stably.
The major feature of the system is self-learning. No separate or pre-training is required. It
can start with an empty database and get to know the faces of the people showing up in an
unsupervised way. When known people enter again, the system can recognize them and
adaptively update the corresponding databases to keep up with recent views. The proposed
system can find promising applications in many fields especially for consumer electronics.

III Acknowledgements
My first and deepest gratitude is to my advisor, Prof. Dr.-Ing. Albrecht Rothermel, for his
invaluable supervision and support. His open-minded way of thinking, continuous
encouragement and trust make me feel confident to create novel research ideas. His depth of
knowledge, brilliant mind and penetrating insight into interdisciplinary research guide my
work to be always kept on the right track.
I am deeply grateful to Dr. Rainer Schweer and his group from Deutsche Thomson-Brandt
GmbH for providing a wide research topic and for giving me a free hand to explore
promising research directions. Moreover, through stimulating discussions, I have been
benefited so much from his extensive industrial research experiences.
I would like to express my warmest thanks to Prof. Dr. Heiko Neumann, for his commitment
to be the second referee for this dissertation. His earnest review and comments are crucial to
shape the thesis.
I am also profoundly indebted to Prof. Dr.-Ing. Hans-Jörg Pfleiderer, the dean of the
Engineering School and the head of the Microelectronics department, for providing a superb
research environment.
My special thanks also go to all members of the video-group: Dr.-Ing. Roland Lares, Martin
Lallinger, Ralf Schreier, Karsten Schmidt, Thomas Kumpf and Christian Günter for their
friendly help and inspirational discussions and comments. I wish to thank Lin Wang and Fei
Fei for their contributions to the software interface implementation.
My deep acknowledges go to the following current and previous colleagues and friends: Dr.-
Ing. Ralf Altherr, Dr.-Ing. Markus Buck, Dr.-Ing. Sviatoslav Bulach, Markus Bschorr, Zhen
Chang, Tiefeng Chen, Shan Chen, Turgut Dogan, Richard Geißler, Roland Hacker, Frank
Hagmeyer, Stefan Hirsch, Cheng Miao, Oliver Pfänder, Ivan Perisa, Markus Prokein, Xavier
Queffelec, Wolfgang Schlecker, Walter Schweigart, Lei Wang, Yi Wang, and Chi Zhang for
their generous help to collect the face databases.
I am so grateful to the department secretary, Ehrentraud Höfer for her excellent
administrative work, to the system administrator Markus Prokein and Walter Schweigart for
their great technical supports.
In particular, I thank my dearest parents for their endless love. I am so lucky as well that,
from childhood, I have been influenced by their strong philosophical background. Without
these emotional and mental supports, the achievement of the thesis could not be possible.


IV Table of Contents
Chapter 1 Introduction ..........................................................................................................1
1.1 Motivation................1
1.2 Proposed Approach..2
1.3 Prospective Applications ..........................................................................................4
1.3.1 Home Security..................................4
1.3.2 Automotive.......5
1.3.3 Entertainment....................................................................5
1.3.4 Mobile Phone....6
1.4 Dissertation Outline..................................6
Chapter 2 Background and Related Research.......7
2.1 Biometric Recognition..............................................................................................7
2.2 Face Detection..........8
2.3 Face Tracking.........................................10
2.4 Face Recognition....11
2.4.1 Overview ........................................................................11
2.4.2 Recognition Procedures and Methods............................13
2.4.3 Video-based Recognition...............................................17
2.4.4 Unsupervised and Automatic Systems...........................19
Chapter 3 Combined Face Detection and Tracking Methods.............25
3.1 Introduction ............................................................................25
3.2 Image-based Face Detection...................................................26
3.2.1 Choice of the Detection Algorithm................................26
3.2.2 Overview of the Detection Algorithm............................26
3.2.3 Face Region Estimation..................................................27
3.2.4 Face Detection Quality...................................................29
3.3 Temporal-based Face Detection.............30
3.3.1 Overview ........................................................................30
3.3.2 Search Region Estimation..............30
3.3.3 Analysis of Temporal Changes......................................34
3.4 Summary.................................................................................37
3.5 Further Discussions................................38
Chapter 4 Automatic Face Recognition..............40
4.1 Overview................................................40
4.2 Feature Extraction and Encoding ...........................................40
4.3 Matching/Classification..........................41
4.3.1 Image-based Classifier...................41
4.3.2 Adaptive Similarity Threshold.......................................44
4.3.3 Temporal Filtering..........................................................................................45
4.4 Combined Same Face Decision Algorithms...........................48
Chapter 5 Unsupervised Face Database Construction........................53
5.1 Overview ................................................................................................................53
5.2 Backgrounds for Constructing Face Databases......................53
5.2.1 Supervised Learning.......................53
5.2.2 Unsupervised Learning...................54
5.2.3 Clustering Analysis ........................................................................................56
5.3 Database Structure..................................57
5.3.1 A fused Clustering Method............57
5.3.2 Parameters in the Proposed Structure.............................62

V
5.4 Features of an Optimum Database..........................................................................65
Chapter 6 State Machine Based Automatic Procedure.......................67
6.1 Overview ................................................67
6.2 States Explorations.................................................................67
Chapter 7 System Implementation ......................................................71
7.1 Overview................................................71
7.2 Hardware Configuration.........................71