Learning in content based image retrieval - a brief review
Yan Gao, Kap Luk Chan, Wei‐Yun Yau · 2007
In recent works on content based image retrieval, machine learning has been playing an even more important role. This is motivated by the need to bridge the semantic gap between the low-level visual features and the high-level human perception. This paper presents a review of the latest development in learning methodologies applied to CBIR. It is reviewed from three aspects: discriminative classification, generative modeling and similarity learning. Through this review, we observe the trends in learning for CBIR and conclude with future research directions.