Title :
Learning temporal context in active object recognition using Bayesian analysis
Author :
Paletta, Lucas ; Rantl, Manfred ; Pinz, Axel
Author_Institution :
Joanneum Res., Inst. of Digital Image Process., Graz, Austria
Abstract :
Active object recognition is a successful strategy to reduce the uncertainty of single view recognition, by planning sequences of views, actively obtaining these views, and integrating multiple recognition results. Understanding recognition as a sequential decision problem challenges the visual agent to select discriminative information sources. The presented system emphasizes the importance of temporal context in disambiguating initial object hypotheses, provides the corresponding theory for Bayesian fusion processes, and demonstrates its performance to be superior to alternative view planning schemes. Instance based learning proposed to estimate the control function enables then real-time processing with improved performance characteristics
Keywords :
Bayes methods; active vision; entropy; learning (artificial intelligence); multi-agent systems; object recognition; planning (artificial intelligence); sensor fusion; Bayesian analysis; Bayesian fusion processes; active object recognition; discriminative information sources; instance based learning; performance characteristics; sequential decision problem; single view recognition; temporal context; visual agent; Bayesian methods; Digital images; Digital signal processing; Electric variables measurement; Image analysis; Image recognition; Object recognition; Process planning; Signal analysis; Strategic planning;
Conference_Titel :
Pattern Recognition, 2000. Proceedings. 15th International Conference on
Conference_Location :
Barcelona
Print_ISBN :
0-7695-0750-6
DOI :
10.1109/ICPR.2000.905482