Title :
Familiar speaker recognition
Author :
Wenndt, Stanley J. ; Mitchell, Ronald L.
Author_Institution :
Air Force Res. Lab., Rome, NY, USA
Abstract :
Speaker recognition by machines can be quite good for large groups as seen in NIST speaker recognition evaluations. However, speaker recognition by machine can be fragile for changing environments. This research examines how robust humans are for recognizing familiar speakers in changing environments. Additionally, bandlimited noise was used to try to learn what frequency regions are important for human listeners to recognize familiar speakers.
Keywords :
speaker recognition; NIST speaker recognition evaluations; bandlimited noise; frequency regions; robust humans; Additive noise; Auditory system; High definition video; Humans; Speaker recognition; Speech; Speaker Familiarity; Voice Recognitions by Humans;
Conference_Titel :
Acoustics, Speech and Signal Processing (ICASSP), 2012 IEEE International Conference on
Conference_Location :
Kyoto
Print_ISBN :
978-1-4673-0045-2
Electronic_ISBN :
1520-6149
DOI :
10.1109/ICASSP.2012.6288854