Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Reduce working memory load for visual classification tasks through gaze-based interaction

: Geisler, J.; Granacher, T.

Preprint urn:nbn:de:0011-n-1519788 (237 KByte PDF)
MD5 Fingerprint: ba197dad593d9be370c002b7c7f7063d
Erstellt am: 10.2.2011

International Federation of Automatic Control -IFAC-; International Federation for Information Processing -IFIP-; International Federation of Operational Research Societies -IFORS-; International Ergonomics Association -IEA-:
11th IFAC/IFIP/IFORS/IEA Symposium on Analysis, Design, and Evaluation of Human-Machine Systems 2010 : Held in Valenciennes, France; August 31- September 3, 2010
Valenciennes, 2010
6 S.
Symposium on Analysis, Design, and Evaluation of Human-Machine Systems <11, 2010, Valenciennes>
Konferenzbeitrag, Elektronische Publikation
Fraunhofer IOSB ()
visual classification; gaze control; scene analysis; working memory; gaze tracking

The rate of working memory errors as an influence on the performance of visual classification at computer screens, e. g. in image exploitation, is expected to get reduced by use of gaze tracking instead of conventional pointing techniques. We compared two gaze-based techniques with the usage of a computer mouse: pure visual fixation (PFix) and visual fixation with confirmation (FixC). A prediction of memory errors while performing visual classification has been carried out using the "Human Processor Modeling Language" with the result that the least errors are expected for PFix followed by FixC and MOUSE. This prediction has been confirmed by empirical validation.