用可学习的有理曲线增强图像对比度，提升光照鲁棒性
GC-ART: Global Learnable Second-Order Rational Tone Curves for Illumination Robustness
- 基于通道软直方图生成有理色调曲线，点对点映射保持边缘位置
- 在对比度损坏下达到48.45%准确率，优于基线和现有方法
- 参数少、计算量低，适合资源受限场景的图像预处理
我们提出GC-ART（全局可学习有理色调映射），一种轻量级可微分预处理模块，用于提升图像分类的光照鲁棒性。GC-ART通过643参数的MLP，从每通道软直方图预测一个端点固定的有理色调曲线，并在分类器前进行点对点映射。模块与交叉熵损失及软单调性惩罚联合训练。在CIFAR-10上使用CIFAR风格的ResNet-18，GC-ART在干净数据上保持基线准确率，对乘性变暗情况表现更优，在对比度损坏下达到48.45%准确率（基线46.27%，Zero-DCE++为47.13%）。结果表明，基于直方图的有理曲线能学习有效的全局色调修正，包括扩展对比度行为，同时通过点对点映射自然保留边缘。相比卷积型可学习增强器，其在32×32输入下所需浮点运算量显著更低。当前超参数未调优，仍有优化空间。
原文摘要 · Abstract (English)
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.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。