The NPU System for the 2020 Personalized Voice Trigger Challenge
Published in
arXiv.org(2021)External Links:
TL
TL;DR
This paper describes the system developed by the NPU team for the 2020 personalized voice trigger challenge, consisting of two independently trained subsystems: a small footprint keyword spotting (KWS) system and a speaker verification (SV) system.
Abstract
This paper describes the system developed by the NPU team for the 2020 personalized voice trigger challenge. Our submitted system consists of two independently trained subsystems: a small footprint keyword spotting (KWS) system and a speaker verification (SV) system. For the KWS system, a multi-scale dilated temporal convolutional (MDTC) network is proposed to detect wake-up word (WuW). For SV system, Write something here. The KWS predicts posterior probabilities of whether an audio utterance contains WuW and estimates the location of WuW at the same time. When the posterior probability ofWuW reaches a predefined threshold, the identity information of triggered segment is determined by the SV system. On evaluation dataset, our submitted system obtains detection costs of 0.081and 0.091 in close talking and far-field tasks, respectively.
References16 items
1
The 2020 Personalized Voice Trigger Challenge: Open Database, Evaluation Metrics and the Baseline Systems
2
Accurate Detection of Wake Word Start and End Using a CNN
3
Mining Effective Negative Training Samples for Keyword Spotting
4
Supervised Contrastive Learning
5