Abstract
Curse of dimensionality is a practical and challengingproblem in image categorization, especially in cases witha large number of classes. Multi-class classification en-counters severe computational and storage problems whendealing with these large scale tasks. In this paper, we propose hierarchical feature hashing to effectively reduce dimensionality of parameter space without sacrificing classification accuracy, and at the same time exploit informationin semantic taxonomy among categories. We provide detailed theoretical analysis on our proposed hashing method. Moreover, experimental results on object recognition and scene classification further demonstrate the effectiveness of hierarchical feature hashing.