ICASSP
2010
IEEE
15 years 2 months ago
2010
IEEE
We propose a new type of audio feature (HFCC-ENS) as well as an unsupervised method for detecting short sequences of spoken words (key-phrases) within long speech recordings. Our ...