Abstract
Many vision tasks require a multi-class classififier to discriminate multiple categories, on the order of hundreds or thousands. In this paper, we propose sparse output coding, a principled way for large-scale multi-class classifification, by turning high-cardinality multi-class categorization into a bit-by-bit decoding problem. Specififically, sparse output coding is composed of two steps: effificient coding matrix learning with scalability to thousands of classes, and probabilistic decoding. Empirical results on object recognition and scene classifification demonstrate the effectiveness of our proposed approach