EdgeSpeechNets: Highly Efficient Deep Neural Networks for Speech Recognition on the Edge

TL;DR

This paper introduces EdgeSpeechNets, a family of highly efficient deep neural networks designed for speech recognition on edge devices, combining human and machine design strategies to achieve high accuracy with minimal resource usage.

Contribution

It presents a novel human-machine collaborative approach for designing low-footprint DNNs, resulting in EdgeSpeechNets that outperform existing models in accuracy and efficiency for on-device speech recognition.

Findings

EdgeSpeechNets achieve ~97% accuracy on Google Speech Commands dataset.

EdgeSpeechNets are up to 7.8x smaller and 36x more computationally efficient.

They enable real-time speech recognition on resource-constrained devices.

Abstract

Despite showing state-of-the-art performance, deep learning for speech recognition remains challenging to deploy in on-device edge scenarios such as mobile and other consumer devices. Recently, there have been greater efforts in the design of small, low-footprint deep neural networks (DNNs) that are more appropriate for edge devices, with much of the focus on design principles for hand-crafting efficient network architectures. In this study, we explore a human-machine collaborative design strategy for building low-footprint DNN architectures for speech recognition through a marriage of human-driven principled network design prototyping and machine-driven design exploration. The efficacy of this design strategy is demonstrated through the design of a family of highly-efficient DNNs (nicknamed EdgeSpeechNets) for limited-vocabulary speech recognition. Experimental results using the Google Speech Commands dataset for limited-vocabulary speech recognition showed that EdgeSpeechNets have higher accuracies than state-of-the-art DNNs (with the best EdgeSpeechNet achieving ~97% accuracy), while achieving significantly smaller network sizes (as much as 7.8x smaller) and lower computational cost (as much as 36x fewer multiply-add operations, 10x lower prediction latency, and 16x smaller memory footprint on a Motorola Moto E phone), making them very well-suited for on-device edge voice interface applications.