Neural ODE with Temporal Convolution and Time Delay Neural Networks for Small-Footprint Keyword Spotting
Published in
arXiv.org(2020)External Links:
TL
TL;DR
This paper presents techniques to apply NODE to KWS that make it possible to adopt Batch Normalization to NODE-based network and to reduce the number of computations during inference.
Abstract
In this paper, we propose neural network models based on the neural ordinary differential equation (NODE) for small-footprint keyword spotting (KWS). We present techniques to apply NODE to KWS that make it possible to adopt Batch Normalization to NODE-based network and to reduce the number of computations during inference. Finally, we show that the number of model parameters of the proposed model is smaller by 68% than that of the conventional KWS model.
References10 items
1
A Time Delay Neural Network with Shared Weight Self-Attention for Small-Footprint Keyword Spotting
2
Temporal Convolution for Real-time Keyword Spotting on Mobile Devices
3
Augmented Neural ODEs
4
Neural Ordinary Differential Equations
5