Show simple item record

A new biologically motivated framework for robust object recognition

dc.date.accessioned2005-12-22T02:15:45Z
dc.date.accessioned2018-11-24T10:24:18Z
dc.date.available2005-12-22T02:15:45Z
dc.date.available2018-11-24T10:24:18Z
dc.date.issued2004-11-14
dc.identifier.urihttp://hdl.handle.net/1721.1/30504
dc.identifier.urihttp://repository.aust.edu.ng/xmlui/handle/1721.1/30504
dc.description.abstractIn this paper, we introduce a novel set of features for robust object recognition, which exhibits outstanding performances on a variety ofobject categories while being capable of learning from only a fewtraining examples. Each element of this set is a complex featureobtained by combining position- and scale-tolerant edge-detectors overneighboring positions and multiple orientations.Our system - motivated by a quantitative model of visual cortex -outperforms state-of-the-art systems on a variety of object imagedatasets from different groups. We also show that our system is ableto learn from very few examples with no prior category knowledge. Thesuccess of the approach is also a suggestive plausibility proof for aclass of feed-forward models of object recognition in cortex. Finally,we conjecture the existence of a universal overcompletedictionary of features that could handle the recognition of all objectcategories.
dc.format.extent10 p.
dc.format.extent17638397 bytes
dc.format.extent793841 bytes
dc.language.isoen_US
dc.subjectAI
dc.subjectvisual cortex
dc.subjectobject recognition
dc.subjectface detection
dc.subjecthierarchy
dc.subjectfeature learning
dc.titleA new biologically motivated framework for robust object recognition


Files in this item

FilesSizeFormatView
MIT-CSAIL-TR-2004-074.pdf793.8Kbapplication/pdfView/Open
MIT-CSAIL-TR-2004-074.ps17.63Mbapplication/postscriptView/Open

This item appears in the following Collection(s)

Show simple item record