用稀疏内部快照实现扩散模型无监督异常检测的公平对比
Backbone-Equated Diffusion OOD via Sparse Internal Snapshots
- 设计统一基准协议，消除不同主干网络和测试成本差异影响
- 仅需少量低噪声状态激活值即可达到顶尖检测性能
- 适合研究扩散模型表征特性或轻量级异常检测应用者
扩散模型的无监督异常检测（OOD）比较困难，因结果受主干网络选择、扰动参数化方式和测试时预算影响。本文提出互洽主干对齐（MBE）协议，统一标准扰动水平与逻辑测试成本，使不同扩散主干间可比。在此框架下，引入标准特征快照（CFS）系列检测器，仅利用冻结扩散主干在典型低噪声水平下的极少量原生内部激活值进行探测。在控制性CIFAR尺度基准上，单次前向传播的CFS(1x2)表现最优，而更小的仅解码器变体仍具竞争力。表明冻结扩散主干揭示的相对异常信号主要集中在少数稀疏内部状态，无需完整去噪轨迹或高容量下游头。进一步通过条件编码-解码互补性、对角分数分离和低噪声扰动稳定性，提供局部诊断理论解释。官方代码已开源：https://github.com/RouzAY/cfs-diffusion-ood/
原文摘要 · Abstract (English)
Fair comparison between diffusion-based OOD detectors is challenging, as conclusions can vary with backbone choice, corruption parameterization, and test-time budget. We address this issue through a Mutualized Backbone-Equated (MBE) protocol that aligns canonical corruption levels and logical test-time cost across diffusion backbones. Within this setting, we introduce Canonical Feature Snapshots (CFS), a family of detectors that probes a frozen diffusion backbone using only a tiny number of native internal activations at canonical low-noise levels. On a controlled CIFAR-scale benchmark, the strongest one-forward CFS variant is CFS(1x2), while an even smaller decoder-only variant remains highly competitive. This shows that much of the relative-OOD signal exposed by frozen diffusion backbones is concentrated in a small number of sparse internal states, rather than requiring full denoising trajectories or high-capacity downstream heads. We further provide a local diagnostic theory explaining these observations through conditional encoder-decoder complementarity, diagonal-score separation, and low-noise corruption stability. The official implementation is available at https://github.com/RouzAY/cfs-diffusion-ood/.
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。