Small-Footprint Keyword Spotting on Raw Audio Data with Sinc-Convolutions (2019-11-05T00:00:00.000000Z)

TL;DR

The end-to-end architecture extracts spectral features using parametrized Sinc-convolutions and achieves the competitive accuracy of 96.4% on Google’s Speech Commands test set with only 62k parameters.

Abstract

Keyword Spotting (KWS) enables speech-based user interaction on smart devices. Always-on and battery-powered application scenarios for smart devices put constraints on hardware resources and power consumption, while also demanding high accuracy as well as real-time capability. Previous architectures first extracted acoustic features and then applied a neural network to classify keyword probabilities, optimizing towards memory footprint and execution time.Compared to previous publications, we took additional steps to reduce power and memory consumption without reducing classification accuracy. Power-consuming audio preprocessing and data transfer steps are eliminated by directly classifying from raw audio. For this, our end-to-end architecture extracts spectral features using parametrized Sinc-convolutions. Its memory footprint is further reduced by grouping depthwise separable convolutions. Our network achieves the competitive accuracy of 96.4% on Google’s Speech Commands test set with only 62k parameters.

Authors

G. Rigoll

11 papers

Simon Mittermaier

1 papers

Ludwig Kurzinger

1 papers

Small-Footprint Keyword Spotting on Raw Audio Data with Sinc-Convolutions

TL;DR

Abstract

Authors

References23 items

Temporal Convolution for Real-time Keyword Spotting on Mobile Devices

Interpretable Convolutional Filters with SincNet

Studying the Effects of Feature Extraction Settings on the Accuracy and Memory Requirements of Neural Networks for Keyword Spotting

Speaker Recognition from Raw Waveform with SincNet

End-to-End Speech Recognition From the Raw Waveform

Speech Commands: A Dataset for Limited-Vocabulary Speech Recognition

Direct modeling of raw audio with DNNS for wake word detection

Hello Edge: Keyword Spotting on Microcontrollers

Learning Filterbanks from Raw Speech for Phone Recognition

Deep Residual Learning for Small-Footprint Keyword Spotting

ShuffleNet: An Extremely Efficient Convolutional Neural Network for Mobile Devices

Depthwise Separable Convolutions for Neural Machine Translation

MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications

Efficient Processing of Deep Neural Networks: A Tutorial and Survey

Xception: Deep Learning with Depthwise Separable Convolutions

Eyeriss: A Spatial Architecture for Energy-Efficient Dataflow for Convolutional Neural Networks

Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift

Adam: A Method for Stochastic Optimization

Efficient object localization using Convolutional Networks

Small-footprint keyword spotting using deep neural networks

1.1 Computing's energy problem (and what we can do about it)

ImageNet classification with deep convolutional neural networks

An Application of Recurrent Neural Networks to Discriminative Keyword Spotting

Field of Study

Journal Information

Name

Page

Venue Information

Name

Type

URL

Alternate Names