提出新方法检测生成模型在医学图像中的性能退化与偏差
Bayesian generative models can flag performance loss, bias, and out-of-distribution image content
- 结合拉普拉斯近似与随机迹估计，实现高效不确定性量化
- 该方法能准确预测重建误差和皮肤镜图像的种族偏差
- 可定位图像中墨水、尺子等分布外内容，适合医疗视觉模型审计
生成模型广泛应用于医学影像任务，如异常检测、特征提取和图像生成。由于基于深度学习，它们对分布偏移敏感，在分布外数据上表现不可靠，可能引发如种族代表性不足等偏差风险。现有不确定性量化方法有限。本文提出SLUG：一种针对变分自编码器（VAE）的新不确定量化方法，融合最新拉普拉斯近似与随机迹估计，可随图像维度平稳扩展。实验表明，该方法的不确定性得分与重建误差及皮肤镜图像中的种族代表性偏差高度相关，且像素级不确定性可有效识别墨水、尺子、斑块等分布外内容，这些内容常导致预测模型产生学习捷径。
原文摘要 · Abstract (English)
Generative models are popular for medical imaging tasks such as anomaly detection, feature extraction, data visualization, or image generation. Since they are parameterized by deep learning models, they are often sensitive to distribution shifts and unreliable when applied to out-of-distribution data, creating a risk of, e.g. underrepresentation bias. This behavior can be flagged using uncertainty quantification methods for generative models, but their availability remains limited. We propose SLUG: A new UQ method for VAEs that combines recent advances in Laplace approximations with stochastic trace estimators to scale gracefully with image dimensionality. We show that our UQ score -- unlike the VAE's encoder variances -- correlates strongly with reconstruction error and racial underrepresentation bias for dermatological images. We also show how pixel-wise uncertainty can detect out-of-distribution image content such as ink, rulers, and patches, which is known to induce learning shortcuts in predictive models.
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。