Multi-Class Image Recognition Based on Relevance Vector Machine
吴慧兰 Wu Huilan, Guodong Liu, Zhaobang Pu · 2009
A new multi-class image recognition method based on relevance vector machine (RVM) and binary tree is proposed. Experiments show that, RVM is a good alternative to the popular support vector machine (SVM), which has comparable classification accuracy to the SVM but with much fewer relevance vectors (RVs) and decision time. Also we designed a novel multi-class method by utilizing both class distances and class distributions. The integrated classification procedure starts with computing all the one-to-rest distances and distributions, and then constructs the binary classifying tree for RVM classification. The multi classification algorithm proposed in this paper performs better than the traditional methods such as One-Against-One, One- Against-Rest, Directed Acyclic Graph and Binary Tree based on class distance both in classification efficiency and classification accuracy.