View source for Publications:Real-Time Face Detection Using Illumination Invariant Features

<div style='display: none'>
== Do not edit this section ==
</div>
{{PublicationSetupTemplate|Author=Klaus Kollreider, Hartwig Fronthaler, Josef Bigun
|PID=239285
|Name=Kollreider, Klaus (klakolx) (Högskolan i Halmstad (2804), Sektionen för Informationsvetenskap, Data– och Elektroteknik (IDE) (3905), Halmstad Embedded and Intelligent Systems Research (EIS) (3938));Fronthaler, Hartwig (Austrian Institute of Technology, Vienna, Austria );Bigun, Josef (josef) (Högskolan i Halmstad (2804), Sektionen för Informationsvetenskap, Data– och Elektroteknik (IDE) (3905), Halmstad Embedded and Intelligent Systems Research (EIS) (3938))
|Title=Real-Time Face Detection Using Illumination Invariant Features
|PublicationType=Conference Paper
|ContentType=Refereegranskat
|Language=eng
|Journal=
|JournalISSN=
|Status=
|Volume=
|Issue=
|HostPublication=Image Analysis : Proceedings
|Conference=15th Scandinavian Conference on Image Analysis, Aalborg, Denmark, June 10-14, 2007
|StartPage=41
|EndPage=50
|Year=2007
|Edition=
|Pages=
|City=Berlin
|Publisher=Springer
|Series=Lecture Notes in Computer Science ; Volume 4522/2007
|SeriesISSN=0302-9743
|ISBN=978-3-540-73039-2
|Urls=
|ISRN=
|DOI=http://dx.doi.org/10.1007/978-3-540-73040-8_5
|ISI=000247364000005
|PMID=
|ScopusId=2-s2.0-38049033993
|NBN=urn:nbn:se:hh:diva-2067
|LocalId=2082/2462
|ArchiveNumber=
|Keywords=Object detection;Image analysis;Biometrics;Direction field;Orientation tensor;Quantized angles;Quangles;AdaBoost
|Categories=Teknik och teknologier (2)
|ResearchSubjects=
|Projects=
|Notes=
|Abstract=<p>A robust object/face detection technique processing every frame in real-time (video-rate) is presented. A methodological novelty are the suggested quantized angle features (“quangles”), being designed for illumination invariance without the need for pre-processing, e.g. histogram equalization. This is achieved by using both the gradient direction and the double angle direction (the structure tensor angle), and by ignoring the magnitude of the gradient. Boosting techniques are applied in a quantized feature space. Separable filtering and the use of lookup tables favor the detection speed. Furthermore, the gradient may then be reused for other tasks as well. A side effect is that the training of effective cascaded classifiers is feasible in very short time, less than 1 hour for data sets of order 104. We present favorable results on face detection, for several public databases (e.g. 93% Detection Rate at 1×10− 6 False Positive Rate on the CMU-MIT frontal face test set).</p>
|Opponents=
|Supervisors=
|Examiners=
|Patent=
|ThesisLevel=
|Credits=
|Programme=
|Subject=
|Uppsok=
|DefencePlace=
|DefenceLanguage=
|DefenceDate=
|CreatedDate=2008-10-20
|PublicationDate=2008-10-20
|LastUpdated=2012-12-10
|diva=http://hh.diva-portal.org/smash/record.jsf?searchId=1&pid=diva2:239285}}
<div style='display: none'>

== Keep all hand-made modifications below ==
</div>{{PublicationDisplayTemplate}}