Abstract

Image retrieval and ranking based on the multi-attribute queries is beneficial to various real world applications. Traditional methods on this prob- lem often utilize intermediate representations generated by attribute classi fiers to describe the images, and then the images in the database are sorted according to their similarities to the query. However, such a scheme has two main chal- lenges: 1) how to exploit the correlation between query attributes and non-query attributes, and 2) how to handle noisy representations since the pre-defined at- tribute classi fiers are probably unreliable. To overcome these challenges, we dis- cover the correlation among attributes via expanding the query representation, and imposing the group sparsity on representations to reduce the disturbance of noisy data. Speci fically, given a multi-attribute query matrix with each row corre- sponding to a query attribute and each column the pre-defined attribute, we firstly expand the query based on the correlation of the attributes learned from the train- ing data. Then, the expanded query matrix is reconstructed by the images in the dataset with the regularization. Furthermore, we introduce the ranking SVM into the objective function to guarantee the ranking consistency. Finally, we adopt a graph regularization to preserve the local visual similarity among images. Ex- tensive experiments on LFW, CUB-200-2011, and Shoes datasets are conducted to demonstrate the effectiveness of our proposed method.