Efficient Keyword Spotting Using Dilated Convolutions and Gating

We explore the application of end-to-end stateless temporal modeling to small-footprint keyword spotting as opposed to recurrent networks that model long-term temporal dependencies using internal states. We propose a model inspired by the recent success of dilated convolutions in sequence modeling a...

Full description

Saved in:
Bibliographic Details
Main Authors: Coucke, Alice, Chlieh, Mohammed, Gisselbrecht, Thibault, Leroy, David, Poumeyrol, Mathieu, Lavril, Thibaut
Format: Conference Proceeding
Language:English
Subjects:
Computational modeling
Computer architecture
Convolution
dilated convolution
end-to-end keyword spotting
Hidden Markov models
Labeling
Logic gates
Noise measurement
open dataset
wake-word detection
Citations: Items that cite this one
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:We explore the application of end-to-end stateless temporal modeling to small-footprint keyword spotting as opposed to recurrent networks that model long-term temporal dependencies using internal states. We propose a model inspired by the recent success of dilated convolutions in sequence modeling applications, allowing to train deeper architectures in resource-constrained configurations. Gated activations and residual connections are also added, following a similar configuration to WaveNet. In addition, we apply a custom target labeling that back-propagates loss from specific frames of interest, therefore yielding higher accuracy and only requiring to detect the end of the keyword. Our experimental results show that our model outperforms a max-pooling loss trained recurrent neural network using LSTM cells, with a significant decrease in false rejection rate. The underlying dataset - "Hey Snips" utterances recorded by over 2.2K different speakers - has been made publicly available to establish an open reference for wake-word detection.
ISSN:2379-190X
DOI:10.1109/ICASSP.2019.8683474