The Larger The Fairer? Small Neural Networks Can Achieve Fairness for Edge Devices

TL;DR

This paper introduces FaHaNa, a neural architecture search framework that finds small, fair, and accurate neural networks suitable for edge devices, addressing fairness concerns in resource-constrained AI applications.

Contribution

FaHaNa is a novel fairness- and hardware-aware neural architecture search method that balances fairness, accuracy, and hardware constraints for edge device deployment.

Findings

FaHaNa identifies neural networks with higher fairness and accuracy.

On dermatology data, FaHaNa outperforms MobileNetV2 in fairness and size.

On Raspberry Pi and Odroid, FaHaNa achieves significant speedups.

Abstract

Along with the progress of AI democratization, neural networks are being deployed more frequently in edge devices for a wide range of applications. Fairness concerns gradually emerge in many applications, such as face recognition and mobile medical. One fundamental question arises: what will be the fairest neural architecture for edge devices? By examining the existing neural networks, we observe that larger networks typically are fairer. But, edge devices call for smaller neural architectures to meet hardware specifications. To address this challenge, this work proposes a novel Fairness- and Hardware-aware Neural architecture search framework, namely FaHaNa. Coupled with a model freezing approach, FaHaNa can efficiently search for neural networks with balanced fairness and accuracy, while guaranteed to meet hardware specifications. Results show that FaHaNa can identify a series of neural networks with higher fairness and accuracy on a dermatology dataset. Target edge devices, FaHaNa finds a neural architecture with slightly higher accuracy, 5.28x smaller size, 15.14% higher fairness score, compared with MobileNetV2; meanwhile, on Raspberry PI and Odroid XU-4, it achieves 5.75x and 5.79x speedup.