SISU - A Speaker Identification System from Short Utterances
Lists
Only you see this
Abstract
Technology has made a paramount impact in our daily life over the last decade by assisting us in ways more than we could have imagined of in the last century. Safeguarding our identity in the digital world has been one of the primary concerns in this era and scientists have devoted their attention to biometric security for the same due to its array of advantages. Humans can be identified using a lot of biometrics and voice is one of them. SISU (Speaker Identification from Short Utterances) is a system proposed towards identification of humans from voice clips of very short length. The system works by Mel Frequency Cepstral Coefficient (MFCC) based features. The system was tested on a short utterance phoneme database of 3290 clips and a highest accuracy of 96.66% was obtained using Random Forest amidst different classifiers for the system.
BibTeX
@inproceedings{mukherjee2018sisu,
author = {Himadri Mukherjee and Moumita Dutta and Sk Md Obaidullah and KC Santosh and Santanu Phadikar and Kaushik Roy},
title = {{SISU - A Speaker Identification System from Short Utterances}},
booktitle = {RTIP2R},
year = {2018},
doi = {10.1007/978-981-13-9181-1\_39},
}Loading related content...