IMPROVING THE EFFICIENCY OF IMAGE RETRIEVAL BASED ON LEARNING THE DISTANCE MEASURE WITH RELEVANCE FEEDBACK

Not scheduled
20m
BHSS. Academia Sinica

BHSS. Academia Sinica

Poster Presentation Track 10: Artificial Intelligence (AI)

Speakers

Dr TRUNG HOANG XUAN (University of Business and Technology, Ministry of Industry and Trade) Dr TUYET DAO VAN (Vietnam National Space Center, Vietnam Academy of Science and Technology)

Description

ABSTRACT
The paper presents an effective method to learn the distance measure between images based on two processing phases: filter high-distinct images with query image and combine the feature sets of images. Our proposed algorithm calculates the non-linear distance measure of images, using the Fuzzy integral and based on the relevance feedback. The experiments have demonstrated that the performance of content-based image retrieval increases significantly when using the proposed measure. In the next study, we will continue the research direction to train and determine the measure on any image.
Keywords: Content-based image retrieval, relevance feedback, fuzzy measure.

Primary authors

Dr TRUNG HOANG XUAN (University of Business and Technology, Ministry of Industry and Trade) Dr TUYET DAO VAN (Vietnam National Space Center, Vietnam Academy of Science and Technology)

Presentation materials

There are no materials yet.