MoENAS: Mixture-of-Expert based Neural Architecture Search for jointly Accurate, Fair, and Robust Edge Deep Neural Networks

TL;DR

MoENAS introduces a neural architecture search method that optimizes edge DNNs for accuracy, fairness, and robustness, significantly reducing disparities and improving performance on skin tone classification tasks.

Contribution

This paper presents MoENAS, a novel automatic design technique using mixture-of-experts to create more accurate, fair, and robust edge DNNs, addressing overlooked metrics in prior methods.

Findings

Reduces skin tone accuracy disparity from 14.09% to 5.60%

Improves overall accuracy by 4.02% over SOTA models

Enhances robustness by 3.80% and minimizes overfitting

Abstract

There has been a surge in optimizing edge Deep Neural Networks (DNNs) for accuracy and efficiency using traditional optimization techniques such as pruning, and more recently, employing automatic design methodologies. However, the focus of these design techniques has often overlooked critical metrics such as fairness, robustness, and generalization. As a result, when evaluating SOTA edge DNNs' performance in image classification using the FACET dataset, we found that they exhibit significant accuracy disparities (14.09%) across 10 different skin tones, alongside issues of non-robustness and poor generalizability. In response to these observations, we introduce Mixture-of-Experts-based Neural Architecture Search (MoENAS), an automatic design technique that navigates through a space of mixture of experts to discover accurate, fair, robust, and general edge DNNs. MoENAS improves the accuracy by 4.02% compared to SOTA edge DNNs and reduces the skin tone accuracy disparities from 14.09% to 5.60%, while enhancing robustness by 3.80% and minimizing overfitting to 0.21%, all while keeping model size close to state-of-the-art models average size (+0.4M). With these improvements, MoENAS establishes a new benchmark for edge DNN design, paving the way for the development of more inclusive and robust edge DNNs.