Query Difficulty Guided Image Retrieval System.

Yangxi Li, Yong Luo, Dacheng Tao, Chao Xu

Abstract: Query difficulty estimation is a useful tool for content-based image retrieval. It predicts the performance of the search result of a given query, and thus it can guide the pseudo relevance feedback to rerank the image search results, and can be used to re-write the given query by suggesting “easy” alternatives. This paper presents a query difficulty estimation guided image retrieval system. The system initially estimates the difficulty of a given query image by analyzing both the query image and the retrieved top ranked images. Different search strategies are correspondingly applied to improve the retrieval performance.