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

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.

Disaster Image Filtering and Summarization Based on Multi-layered Affinity Propagation

“…Xie et al [5] propose a Kway min-max cut clustering algorithm for filtering out junk images for Google Image search results. The limitation is that the number of clusters has to be preset, which lacks flexibility and may not match the semantic distribution for an image topic.…”

Exploring Hidden Coherent Feature Groups and Temporal Semantics for Multimedia Big Data Analysis

“…As for image filtering, it involves filtering out irrelevant images returned from typical keyword-based search engines because of mis-correspondence between the keyword and the underlying image semantic. Xie et al [104] propose a K-way min-max cut clustering algorithm for filtering out junk images for Google Image search results, and the work is further extended to inspect the cluster correlations between two different search engines [105]. An inherited limitation with these two approaches is the number of clusters, i.e., K, has to be preset, which lacks the flexibility and may not match the semantic distribution for an image topic.…”