机器人双耳定位信息提取与重构

Extraction and Reconstruction of Robotic Binaural Location Information

机器人的定位技术是机器人研究领域中的重要技术,得到了广泛的研究。针对机器人的双耳定位信息,提出了基于非负矩阵分解和支持向量机回归的方法。首先对影响机器人双耳定位信息的参数进行分析,提取其中的主要参数,其次利用非负矩阵分解对头相关传输函数数据集进行数据降维,以减小数据运算量,最后采用基于统计的非线性回归的方法对头相关传输函数数据集进行任意方位信息的重构。实验表明,所提方法取得了较好的效果。

The location of robots is an important technology in the field of robot research and receives extensive attention. Aiming at the binaural location information of the robot, a method based on non-negative matrix factorization and support vector machine regression is proposed. Firstly, the parameters affecting the binaural location information of the robot are analyzed, and the main parameters extracted. Then, non- negative matrix decomposition is used to perform data dimensionality reduction on the head related transfer function data set and decrease the amount of data operations. Finally, a statistical-based nonlinear regression method is used to reconstruct the orientation information of the head-related transfer function data set. Experiments indicate that the proposed method could achieve fairly good results.