Nonlinear variants of biased discriminants for interactive image retrieval

During an interactive image retrieval process with relevance feedback, kernel-based or boosted learning algorithms can provide superior nonlinear modelling capability. In the paper, such nonlinear extensions for biased discriminants, or BiasMap are discussed. Kernel partial alignment is proposed as the criterion for kernel selection. The associated analysis also provides a gauge on relative class scatters, which can guide an asymmetric learner, such as BiasMap, toward better class modelling. Two boosted versions of BiasMap are also proposed. Unlike existing approaches that boost feature components or vectors to form a composite classifier, the new scheme boosts linear BiasMap toward a nonlinear ranker which is more suited for small-sample learning during interactive image retrieval. Experiments on heterogeneous image database retrieval in addition to small sample face retrieval are used for performance evaluations.