?url_ver=Z39.88-2004&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Adc&rft.title=Learning+image+components+for+object+recognition&rft.creator=Spratling%2C+Michael+W&rft.subject=Machine+Vision&rft.subject=Machine+Learning&rft.subject=Neural+Nets&rft.description=In+order+to+perform+object+recognition+it+is+necessary+to+learn+representations+of+the+underlying+components+of+images.++Such+components+correspond+to+objects%2C+object-parts%2C+or+features.++Non-negative+matrix+factorisation+is+a+generative+model+that+has+been+specifically+proposed+for+finding+such+meaningful+representations+of+image+data%2C+through+the+use+of+non-negativity+constraints+on+the+factors.++This+article+reports+on+an+empirical+investigation+of+the+performance+of+non-negative+matrix+factorisation+algorithms.+It+is+found+that+such+algorithms+need+to+impose+additional+constraints+on+the+sparseness+of+the+factors+in+order+to+successfully+deal+with+occlusion.+However%2C+these+constraints+can+themselves+result+in+these+algorithms+failing+to+identify+image+components+under+certain+conditions.+In+contrast%2C+a+recognition+model+(a+competitive+learning+neural+network+algorithm)+reliably+and+accurately+learns+representations+of+elementary+image+features+without+such+constraints.%0A&rft.date=2006&rft.type=Journal+(Paginated)&rft.type=PeerReviewed&rft.format=application%2Fpdf&rft.identifier=http%3A%2F%2Fcogprints.org%2F4886%2F1%2Fjmlr06.pdf&rft.identifier=++Spratling%2C+Michael+W++(2006)+Learning+image+components+for+object+recognition.++%5BJournal+(Paginated)%5D+++++&rft.relation=http%3A%2F%2Fcogprints.org%2F4886%2F