Face Retrieval with Relevance Feedback Using Lifting Wavelets Features

摘要: By using support vector machine (SVM), this paper presents a novel face retrieval scheme in database based on lifting wavelets features. The relevance feedback mechanism is also performed. can be described three stages as follows. First, decomposition technique employed because it not only extract the optimal intrinsic features for representing image, but accelerate speed of transform. Second, Linear Discriminant Analysis (LDA) adopted to reduce feature dimensionality and enhance class discriminability. Third, SVM applied learn user’s refine performance. experimental evaluation has been conducted ORL dataset which results show that our proposed approach effective promising.