Face Recognition across Large Viewpoint Changes
Author(s) / Creator(s)
O´Toole, Alice J.
Bülthoff, Heinrich H.
Troje, Nikolaus F.
Vetter, Thomas
Abstract / Description
We describe a computational model of face recognition that makes use of the overlapping texture and shape information visible in different views of faces. The model operates on view dependent data from three-dimensional laser scans of human heads, wich provided three-dimensional surface data as well as surface image detail in form of a texture map. View-dependent information from the surface and texture representations was registered onto separate three-dimensional head models. We used an auto-associative memory model as a pattern completion device to fill in parts of the head from a lerned view when a test view with partially overlapping information was used as a memory key- We show that the overlapping visible regions of heads for both surface and texture data can support accurate recognition, even with pose differences of as much as 90 degrees (full face to profile view) between the learning and test view.
Keyword(s)
Gesicht Biometrie Gedächtnis Lernen Gesichtserkennung Drei-dimensionales Kopfmodell Gedächtnis Lernen face recognition three-dimensional head models memory pose differences learningPersistent Identifier
Date of first publication
1994
Citation
-
TR_009.pdfAdobe PDF - 276.32KBMD5: 78417933a2c016cf3ad6f0a2b5038e8e
-
There are no other versions of this object.
-
Author(s) / Creator(s)O´Toole, Alice J.
-
Author(s) / Creator(s)Bülthoff, Heinrich H.
-
Author(s) / Creator(s)Troje, Nikolaus F.
-
Author(s) / Creator(s)Vetter, Thomas
-
PsychArchives acquisition timestamp2022-11-17T11:03:35Z
-
Made available on2007-03-14
-
Made available on2015-12-01T10:32:05Z
-
Made available on2022-11-17T11:03:35Z
-
Date of first publication1994
-
Abstract / DescriptionWe describe a computational model of face recognition that makes use of the overlapping texture and shape information visible in different views of faces. The model operates on view dependent data from three-dimensional laser scans of human heads, wich provided three-dimensional surface data as well as surface image detail in form of a texture map. View-dependent information from the surface and texture representations was registered onto separate three-dimensional head models. We used an auto-associative memory model as a pattern completion device to fill in parts of the head from a lerned view when a test view with partially overlapping information was used as a memory key- We show that the overlapping visible regions of heads for both surface and texture data can support accurate recognition, even with pose differences of as much as 90 degrees (full face to profile view) between the learning and test view.en
-
Persistent Identifierhttps://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:bsz:291-psydok-9290
-
Persistent Identifierhttps://hdl.handle.net/20.500.11780/1044
-
Persistent Identifierhttps://doi.org/10.23668/psycharchives.9002
-
Language of contenteng
-
Keyword(s)Gesichtde
-
Keyword(s)Biometriede
-
Keyword(s)Gedächtnisde
-
Keyword(s)Lernende
-
Keyword(s)Gesichtserkennungde
-
Keyword(s)Drei-dimensionales Kopfmodellde
-
Keyword(s)Gedächtnisde
-
Keyword(s)Lernende
-
Keyword(s)face recognitionen
-
Keyword(s)three-dimensional head modelsen
-
Keyword(s)memoryen
-
Keyword(s)pose differencesen
-
Keyword(s)learningen
-
Dewey Decimal Classification number(s)150
-
TitleFace Recognition across Large Viewpoint Changesen
-
DRO typereport
-
Visible tag(s)PsyDok