Mean Hilbert envelope coefficients (MHEC) for robust speaker and language identification

Speech Communication（2015）

引用 72|浏览38

暂无评分

摘要

•Auditory robust feature for speaker and language identification (SID/LID).•Employs Hilbert envelope of Gammatone filterbank outputs.•Excellent performance in extreme noisy/channel conditions (DARPA RATS data).•Power-law non-linearity gives consistent improvement for SID/LID tasks.

查看译文

关键词

Language identification,MHEC,Mismatch conditions,Robust features,Speaker identification

AI 理解论文

溯源树

样例

生成溯源树，研究论文发展脉络

Chat Paper

正在生成论文摘要