发表机构
The Hong Kong University of Science and Technology（香港科技大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
研究图像去噪中暗区域信噪比低、MSE训练去噪器加剧偏差的问题，提出亮度偏差鲁棒去噪（BBRD）方法，将像素分亮度带，归一化误差并应用Group-DRO，实验证明该方法能同时改善各亮度带，在暗区域增益最大。
AI 中文摘要
本文揭示了图像去噪中一个重要却被忽视的问题：在信号依赖的相机噪声模型下，暗区域的信噪比（SNR）天生较低，信号强度衰减比噪声方差减小快得多，使得暗区域细节恢复极具挑战。然而，基于均方误差（MSE）训练的去噪器非但没有弥补这一困难，反而加剧了它——重构暗像素比其每波段噪声本底差6倍。这种偏差源于两个因素：信号依赖噪声使亮像素残差膨胀，且网络的雅可比范数随亮度单调增加。为此，我们提出亮度偏差鲁棒去噪（BBRD），它是MSE损失的替代方法，将像素划分为亮度带，通过经验噪声方差对每带误差进行归一化，并应用组分布鲁棒优化（Group-DRO）动态加重当前最差的带，无需额外参数或推理成本。实验表明，BBRD是13种测试方法中唯一能同时改善各亮度带的方法，在暗带可达+0.45 dB，亮带可达+0.32 dB，在SIDD数据集上总峰值信噪比（PSNR）可达+0.65 dB，在最暗区域增益最大。代码可从该https链接获取。
英文摘要
In this paper, we reveal an important yet overlooked problem in image denoising: under signal-dependent camera noise models, dark regions suffer from inherently low Signal-to-Noise Ratio (SNR), as signal intensity decays far faster than noise variance diminishes, making detail recovery in dark areas fundamentally challenging. Yet rather than compensating for this difficulty, MSE-trained denoisers exacerbate it -- reconstructing dark pixels up to 6x worse relative to their per-band noise floor. This bias stems from two compounding factors: signal-dependent noise inflates bright-pixel residuals, and the network's Jacobian norm increases monotonically with brightness. Together, these cause bright regions to chronically dominate gradient updates at the expense of dark ones. To this end, we propose Brightness Bias-Robust Denoising (BBRD), a drop-in replacement for MSE loss that partitions pixels into brightness bands, normalizes per-band error by empirical noise variance, and applies Group Distributionally Robust Optimization (Group-DRO) to dynamically upweight whichever band is currently worst, with zero additional parameters or inference cost. Across 8 architectures and 2 datasets in our experiments, BBRD is the only method among 13 tested alternatives that improves each brightness band simultaneously, achieving up to +0.45 dB on dark bands, +0.32 dB on bright bands, and +0.65 dB aggregate Peak Signal-to-Noise Ratio (PSNR) on SIDD, with the largest per-band gains in the darkest regions where detail recovery matters most. Code is available at https://github.com/xmed-lab/BBRD
Comments6 figures, 4 tables. Code: https://github.com/xmed-lab/BBRD
