CASCADE OF MULTI-LEVEL MULTI-INSTANCE CLASSIFIERS FOR IMAGE ANNOTATION

Cam-Tu Nguyen, Ha Vu Le, Takeshi Tokuyama

2011

Abstract

This paper introduces a new scheme for automatic image annotation based on cascading multi-level multiinstance classifiers (CMLMI). The proposed scheme employs a hierarchy for visual feature extraction, in which the feature set includes features extracted from the whole image at the coarsest level and from the overlapping sub-regions at finer levels. Multi-instance learning (MIL) is used to learn the “weak classifiers” for these levels in a cascade manner. The underlying idea is that the coarse levels are suitable for background labels such as “forest” and “city”, while finer levels bring useful information about foreground objects like “tiger” and “car”. The cascade manner allows this scheme to incorporate “important” negative samples during the learning process, hence reducing the “weakly labeling” problem by excluding ambiguous background labels associated with the negative samples. Experiments show that the CMLMI achieve significant improvements over baseline methods as well as existing MIL-based methods. improvements over baseline methods as well as existing MIL-based methods.

Paper Citation

in Harvard Style

Nguyen C., Le H. and Tokuyama T. (2011). CASCADE OF MULTI-LEVEL MULTI-INSTANCE CLASSIFIERS FOR IMAGE ANNOTATION . In Proceedings of the International Conference on Knowledge Discovery and Information Retrieval - Volume 1: KDIR, (IC3K 2011) ISBN 978-989-8425-79-9, pages 14-23. DOI: 10.5220/0003634400140023

in Bibtex Style

@conference{kdir11,
author={Cam-Tu Nguyen and Ha Vu Le and Takeshi Tokuyama},
title={CASCADE OF MULTI-LEVEL MULTI-INSTANCE CLASSIFIERS FOR IMAGE ANNOTATION},
booktitle={Proceedings of the International Conference on Knowledge Discovery and Information Retrieval - Volume 1: KDIR, (IC3K 2011)},
year={2011},
pages={14-23},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0003634400140023},
isbn={978-989-8425-79-9},
}

in EndNote Style

TY - CONF
JO - Proceedings of the International Conference on Knowledge Discovery and Information Retrieval - Volume 1: KDIR, (IC3K 2011)
TI - CASCADE OF MULTI-LEVEL MULTI-INSTANCE CLASSIFIERS FOR IMAGE ANNOTATION
SN - 978-989-8425-79-9
AU - Nguyen C.
AU - Le H.
AU - Tokuyama T.
PY - 2011
SP - 14
EP - 23
DO - 10.5220/0003634400140023