| 基于数据生成的手语识别自适应方法 |
| |
| 引用本文： | 周宇,陈熙霖,赵德斌,姚鸿勋,高文. 基于数据生成的手语识别自适应方法[J]. 高技术通讯, 2009, 19(12). DOI: 10.3772/j.issn.1002-0470.2009.12.009 |
| |
| 作者姓名： | 周宇 陈熙霖 赵德斌 姚鸿勋 高文 |
| |
| 作者单位： | 哈尔滨工业大学计算机科学与技术学院,哈尔滨,150001;中国科学院智能信息处理重点实验室,中国科学院计算技术研究所,北京,100190;北京大学信息科学技术学院,北京,100871 |
| |
| 基金项目： | 863计划，北京市自然科学基金，国家自然科学基金 |
| |
| 摘 要： | 采用自适应技术来解决非特定人手语识别问题,提出了一种基于数据生成的手语识别自适应方法.首先,对非特定人模型的均值向量进行自动聚类,寻找聚类中心生成手语词码本,然后,根据此码本选择词根子集,该子集能覆盖手语词码本的所有码字,继而,使用新用户的词根子集数据通过遗传算法生成其它词根的数据,最后,结合词根子集的真实数据和其它词根的生成数据,利用最大似然线性回归(MLLR)和最大后验概率(MAP)算法对非特定人模型进行自适应.实验结果表明,该方法既能够降低所需要的自适应数据量,又能够在非特定人模型基础上取得识别正确率的大幅提高.
|
| 关 键 词： | 手语识别 自适应 最大似然线性回归(MLLR) 最大后验概率(MAP) 遗传算法 |
An adaptive Chinese sign language recognition method based on data generating |
| |
| Zhou Yu,Chen Xilin,Zhao Debin,Yao Hongxun,Gao Wen. An adaptive Chinese sign language recognition method based on data generating[J]. High Technology Letters, 2009, 19(12). DOI: 10.3772/j.issn.1002-0470.2009.12.009 |
| |
| Authors: | Zhou Yu Chen Xilin Zhao Debin Yao Hongxun Gao Wen |
| |
| Affiliation: | Zhou Yu~*,Chen Xilin~(**),Zhao Debin~*,Yao Hongxun~*,Gao Wen~(***) ( * School of Computer Science and Technology,Harbin Institute of Technology,Harbin 150001 ) (** Key Laboratory of Intelligent Information Processing,Chinese Academy of Sciences,Institute of Computing Technology,Beijing 100190) (***School of Electronics Engineering and Computer Science,Peking University,Beijing 100871 ) |
| |
| Abstract: | This paper proposes an adaptive Chinese sign language recognition method based on data generating to solve the signer-independent (SI) problem using the adaptation technology. The method is described as below. First, the SI models' means are automatic clustered, and then the cluster centers are taken as codebooks of all etyma. A subset of all etyma is selected to cover all codewords. The data of the subset for a new signer are used to generate the data of other etyma with the genetic algorithm. By utilizing... |
| |
| Keywords: | sign language recognition adaptation Maximum Likelihood Linear Regression (MLLR) Maximum a Posteriori (MAP) genetic algorithm |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
