GC-ART: Global Learnable Second-Order Rational Tone Curves for Illumination Robustness

TL;DR

GC-ART is a lightweight, differentiable pre-processing module that learns global tone curves for robust image classification, improving performance on contrast and darkening corruptions while maintaining efficiency.

Contribution

It introduces a novel global rational tone curve predictor conditioned on histograms, enhancing robustness with fewer FLOPs compared to convolutional methods.

Findings

Matches clean accuracy on CIFAR-10 with baseline

Improves robustness on darkening and contrast corruptions

Uses significantly fewer FLOPs than convolutional enhancers

Abstract

We introduce GC-ART (Global Curve Adaptive Rational Tone-mapping), a lightweight differentiable pre-processing module for robust image classification. GC-ART predicts an endpoint-pinned rational tone curve from per-channel soft histograms using a 643-parameter MLP, then applies the curve pointwise before the classifier. The module is trained end-to-end with cross-entropy and a soft monotonicity penalty. On CIFAR-10 with a CIFAR-style ResNet-18, GC-ART matches clean accuracy with the unenhanced baseline and other learned enhancers, improves over the baseline on multiplicative darkening, and achieves the best learned-method result on contrast corruption (48.45% vs. 46.27% for the baseline and 47.13% for Zero-DCE++). These results suggest that histogram-conditioned rational curves can learn useful global tone corrections, including contrast-expanding behavior, while preserving edge locations by construction through pointwise mapping. GC-ART also uses substantially fewer FLOPs than convolutional learned enhancers at 32 x 32. The current hyperparameters are untuned, leaving room for systematic improvement.