MR-MIL: Manifold Ranking Based Multiple-Instance Learning for Automatic Image Annotation

Yufeng ZHAO  Yao ZHAO  Zhenfeng ZHU  Jeng-Shyang PAN  

IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences   Vol.E91-A   No.10   pp.3088-3089
Publication Date: 2008/10/01
Online ISSN: 1745-1337
DOI: 10.1093/ietfec/e91-a.10.3088
Print ISSN: 0916-8508
Type of Manuscript: LETTER
Category: Image
automatic image annotation,  multiple-instance leaning,  manifold ranking,  SVM,  

Full Text: PDF>>
Buy this Article

A novel automatic image annotation (AIA) scheme is proposed based on multiple-instance learning (MIL). For a given concept, manifold ranking (MR) is first employed to MIL (referred as MR-MIL) for effectively mining the positive instances (i.e. regions in images) embedded in the positive bags (i.e. images). With the mined positive instances, the semantic model of the concept is built by the probabilistic output of SVM classifier. The experimental results reveal that high annotation accuracy can be achieved at region-level.