FairQuant: Fairness-Aware Mixed-Precision Quantization for Medical Image Classification

TL;DR

FairQuant introduces a fairness-aware mixed-precision quantization framework for medical image classification, optimizing model size, accuracy, and fairness across groups under explicit bit budgets.

Contribution

It proposes a novel approach combining group-aware importance analysis and learnable bit allocation to improve fairness in quantized models.

Findings

Recover 4-6 bit accuracy close to 8-bit models

Improve worst-group performance over uniform quantization

Maintain fairness metrics under shared bit budgets

Abstract

Compressing neural networks by quantizing model parameters offers useful trade-off between performance and efficiency. Methods like quantization-aware training and post-training quantization strive to maintain the downstream performance of compressed models compared to the full precision models. However, these techniques do not explicitly consider the impact on algorithmic fairness. In this work, we study fairness-aware mixed-precision quantization schemes for medical image classification under explicit bit budgets. We introduce FairQuant, a framework that combines group-aware importance analysis, budgeted mixed-precision allocation, and a learnable Bit-Aware Quantization (BAQ) mode that jointly optimizes weights and per-unit bit allocations under bitrate and fairness regularization. We evaluate the method on Fitzpatrick17k and ISIC2019 across ResNet18/50, DeiT-Tiny, and TinyViT. Results show that FairQuant configurations with average precision near 4-6 bits recover much of the Uniform 8-bit accuracy while improving worst-group performance relative to Uniform 4- and 8-bit baselines, with comparable fairness metrics under shared budgets.