K-way min-max cut for image clustering and junk images filtering from Google images

Authors:
Feng Xie;Yi Shen;Xiaofei He
Affiliations:
State Key Lab of CAD&CD / Zhejiang University, Hangzhou, China;UNC-Charlotte, Charlotte, NC, USA;State Key Lab of CAD&CD / Zhejiang University, Hangzhou, China
Venue:
Proceedings of the international conference on Multimedia
Year:
2010

Citing 6
Cited 0

The Earth Mover's Distance as a Metric for Image Retrieval

International Journal of Computer Vision
Texture Features and Learning Similarity

CVPR '96 Proceedings of the 1996 Conference on Computer Vision and Pattern Recognition (CVPR '96)
Distinctive Image Features from Scale-Invariant Keypoints

International Journal of Computer Vision
Learning Object Categories from Google"s Image Search

ICCV '05 Proceedings of the Tenth IEEE International Conference on Computer Vision - Volume 2
An efficient color representation for image retrieval

IEEE Transactions on Image Processing
Integrating Concept Ontology and Multitask Learning to Achieve More Effective Classifier Training for Multilevel Image Annotation

IEEE Transactions on Image Processing

Quantified Score

Hi-index	0.01

Visualization

Abstract

Currently most existing image search engines such as Google Images index web images majorly using text keywords extracted from the context, which may return large amount of junk information. We propose a novel clustering based filtering method to filter those junk images. Firstly we apply K-way min-max cut to cluster images returned by Google into multiple clusters based on the mixture of feature kernels, with kernel weights being determined automatically instead of hard fix. Secondly we select the best cluster in a robust way, and rank all the rest clusters according to their similarity with the best one. Finally those low-rank clusters can be filtered out as junk clusters. In experiments we obtain very comparative filtering performance to the current state-of-the-art, and improve Google Images search results significantly.