Biologically-inspired object recognition system for recognizing natural scene categories

Ali Alameer, Patrick Degenaar, Kianoush Nazarpour

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Visual processing has attracted a lot of attention in the last decade. Hierarchical approaches for object recognition are gradually becoming widely-accepted. Generally, they are inspired by the ventral stream of human visual cortex, which is in charge of rapid categorization. Similar to objects, natural scenes share common features and can, therefore, be classified in the same manner. However, natural scenes generally show a high level of statistical correlation between classes. This, in fact, is a major challenge for most object recognition models. Rapid categorization of a natural scene in the absence of attention is a challenge. However, researchers have found that 150 ms is enough to categorize a complex natural scene. We tested the capability of our recent and bio-inspired En-HMAX model of visual processing for scene classification. The results show the En-HMAX model has a comparable performance to state of the art methods for natural scene categorization.
Original languageEnglish
Title of host publication2016 International Conference for Students on Applied Engineering (ICSAE)
PublisherInstitute of Electrical and Electronics Engineers (IEEE)
Pages129-132
Number of pages4
ISBN (Electronic)978-1-4673-9053-8
ISBN (Print)978-1-4673-9028-6
DOIs
Publication statusPublished - 9 Jan 2017
Event1st International Conference for Students on Applied Engineering, ICSAE 2016 - Newcastle Upon Tyne, United Kingdom
Duration: 20 Oct 201621 Oct 2016

Conference

Conference1st International Conference for Students on Applied Engineering, ICSAE 2016
Country/TerritoryUnited Kingdom
CityNewcastle Upon Tyne
Period20/10/1621/10/16

Fingerprint

Dive into the research topics of 'Biologically-inspired object recognition system for recognizing natural scene categories'. Together they form a unique fingerprint.

Cite this