Multi-feature fusion algorithm based on generalized discriminative Multi-Set Canonical Correlation Analysis and its application for recognition
Yihai Liu, Jiazhou He, Chun-Shan Ding · 2017
Feature fusion plays an important role in target recognition, especially when single sensor's recognition capability is limited under severe situations. In view of shortcomings of Multi-set Canonical Correlation Analysis (MCCA) and its supervised modified methods in using category information in fusion projection rule learning, a generalized discriminative learning version of MCCA, termed as GDMCCA, is proposed in this paper. GDMCCA can find a set of optimal projection vector for feature fusion such that the fused feature set can simultaneously maximize the difference between within-class and between-class correlation and minimize the total samples' within-class scatter. Results show that the proposed algorithm is more effective and robust than other related feature fusion methods under the same fusion recognition scenarios.