Speech Separation Based on Robust Independent Component Analysis

• 作 者: YAO Wen-po WU Min LIU Tie-bing
• 机 构: NanjingGeneralHospitalofNanjingMilitaryCommand,ImageProcessingandImageCommunicationsKeyLaboratory
• 出 处: 中国生物医学工程学报：英文版
• 发表日期: 2013年4期
• 关键词: 语音分离 独立分量分析 鲁棒性 FASTICA算法 分离试验 语音信号 分离时间 高稳定性 RobustlCA speech separation Fast
• 基金项目: National Natural Science Foundation of China,grant number:61271082,61201029,61102094

In this paper,we applied RobustICA to speech separation and made a comprehensive comparison to FastICA according to the separation results.Through a series of speech signal separation test,RobustICA reduced the separation time consumed by FastICA with hig