Feature Extraction and Evolution Based Pattern Recognition

摘要

This paper proposes a novel method of classifier selection for efficient object recognition based on evolutionary computation and data context knowledge called Evolvable Classifier Selection. The proposed method tries to distinguish the data characteristics of input image (data contexts) and selects a classifier system accordingly using the genetic algorithm. It stores its experiences in terms of the data context category and the artificial chromosome so that the context knowledge can be accumulated and used later. The proposed method operates in two modes: the evolution mode and the action mode. It can improve its performance incrementally using GA in the evolution mode. Once sufficient context knowledge is accumulated, the method can operate in realtime. The proposed method has been evaluated in the area of face recognition. Data context-awareness, modeling and identification of input data as data context categories, is carried out using SOM.