在线草图识别中的用户适应性研究  

Study on User Adaptation for On-Line Sketch Recognition

作　　者：孙正兴[1,2] 彭彬彬[1,2] 丛兰兰 孙建勇[1,2] 张斌 

机构地区：[1]南京大学计算机软件新技术国家重点实验室,南京210093 [2]南京大学计算机科学与技术系,南京210093

出　　处：《计算机辅助设计与图形学学报》2004年第9期1207-1215,共9页Journal of Computer-Aided Design & Computer Graphics

基　　金：国家自然科学基金 ( 6990 3 0 0 6;60 3 73 0 65 )资助

摘　　要：提出一种在线草图识别用户适应性解决方法 ,该方法分别采用支撑向量机主动式增量学习和动态用户建模技术进行笔划和复杂图形的识别 支撑向量机主动式增量学习方法通过主动“分析”用户增量数据 ,并根据用户反馈从中选择重要数据作为训练样本 ,可有效地鉴别用户手绘笔划特征 ,快速地识别用户输入笔划 动态用户建模技术则采用增量决策树记录草图的笔划构成及其手绘过程 ,有效捕捉用户的复杂图形手绘习惯 ,进而利用模糊匹配在草图绘制过程中预测和识别复杂图形 实验表明 :该方法具有很好的效果 。An attempt for user adaptation in online sketch recognition is made. Two important techniques are presented: SVM-based active incremental learning for strokes classification and dynamic user modeling for composite sketchy shape recognition. Firstly, an SVM-based incremental active learning method is performed to adapt users' parameters of strokes. It can actively analyze users' incremental data and make full use of the historical data to reduce the workload of manual labeling and the training time of the classifier. Secondly, a dynamic user modeling method is adopted to build user models for each specific user in an incremental decision tree and recognize dynamically the possible shapes by means of fuzzy matching. Experiments show that the proposed method is effective and efficient.

关 键 词：在线草图识别 用户适应性 主动式增量学习 支撑向量机 动态用户建模 建模技术 人机交互 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] TP18[自动化与计算机技术—计算机科学与技术]