Environmental Sound Classification on the Edge: A Pipeline for Deep Acoustic Networks on Extremely Resource-Constrained Devices

TL;DR

This paper introduces ACDNet, a compact deep neural network for acoustic classification on resource-limited edge devices, achieved through a generic compression pipeline that maintains high accuracy while drastically reducing size and computational requirements.

Contribution

The paper presents a novel, automated pipeline for compressing deep acoustic networks, enabling state-of-the-art performance on extremely resource-constrained edge devices.

Findings

Achieved over 96% accuracy on multiple sound classification datasets.

Reduced network size by 97.22% and FLOPs by 97.28% without significant accuracy loss.

Successfully implemented on a standard microcontroller for real-world applications.

Abstract

Significant efforts are being invested to bring state-of-the-art classification and recognition to edge devices with extreme resource constraints (memory, speed, and lack of GPU support). Here, we demonstrate the first deep network for acoustic recognition that is small, flexible and compression-friendly yet achieves state-of-the-art performance for raw audio classification. Rather than handcrafting a once-off solution, we present a generic pipeline that automatically converts a large deep convolutional network via compression and quantization into a network for resource-impoverished edge devices. After introducing ACDNet, which produces above state-of-the-art accuracy on ESC-10 (96.65%), ESC-50 (87.10%), UrbanSound8K (84.45%) and AudioEvent (92.57%), we describe the compression pipeline and show that it allows us to achieve 97.22% size reduction and 97.28% FLOP reduction while maintaining close to state-of-the-art accuracy 96.25%, 83.65%, 78.27% and 89.69% on these datasets. We describe a successful implementation on a standard off-the-shelf microcontroller and, beyond laboratory benchmarks, report successful tests on real-world datasets.