IncreFA：打破生成模型归属的静态壁垒
IncreFA: Breaking the Static Wall of Generative Model Attribution
把生成图像归属（判别图像来自哪个生成模型）建模为结构化增量学习问题，用架构层级正交先验加潜变量记忆回放来对抗表征漂移；做深伪检测、生成内容溯源或开放集识别的视觉研究者可直接复用其层级约束与伪未见样本混合策略。
原文依据 IncreFA achieves state-of-the-art attribution accuracy and 98.9% unseen detection under a temporally ordered open-set protocol.
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论随着 AI 生成模型以前所未有的速度演进，图像归属已成为一个不断移动的目标。新的扩散、对抗与自回归生成器几乎每月都会出现，使现有的水印、分类器与反演方法一经发布便告过时。核心问题不在于模型识别，而在于归属本身无法自适应。我们提出 IncreFA，该框架将归属重新定义为一个结构化的增量学习问题，使系统能够在新生成模型出现时持续学习。IncreFA 不同于常规增量学习，它利用生成架构之间的层级关系，并将其与持续自适应相耦合。它集成了两个相互强化的机制：（1）层级约束，通过可学习的正交先验编码架构层级，以将族级不变量与模型特有的个体特性解耦；（2）潜变量记忆库，回放紧凑的潜变量样例并对其进行混合以生成伪未见样本，从而稳定表征漂移并增强开放集感知能力。在新建的增量归属基准（IABench）上，该基准覆盖 2022 至 2025 年间发布的 28 个生成模型，IncreFA 在按时间排序的开放集协议下取得了最优的归属准确率与 98.9% 的未见样本检测率。
英文原文摘要 1393 字符
As AI generative models evolve at unprecedented speed, image attribution has become a moving target. New diffusion, adversarial and autoregressive generators appear almost monthly, making existing watermark, classifier and inversion methods obsolete upon release. The core problem lies not in model recognition, but in the inability to adapt attribution itself.We introduce IncreFA, a framework that redefines attribution as a structured incremental learning problem, allowing the system to learn continuously as new generative models emerge. IncreFA departs from conventional incremental learning by exploiting the hierarchical relationships among generative architectures and coupling them with continual adaptation. It integrates two mutually reinforcing mechanisms: (1) Hierarchical Constraints, which encode architectural hierarchies through learnable orthogonal priors to disentangle family-level invariants from model-specific idiosyncrasies; and (2) a Latent Memory Bank, which replays compact latent exemplars and mixes them to generate pseudo-unseen samples, stabilising representation drift and enhancing open-set awareness.On the newly constructed Incremental Attribution Benchmark (IABench) covering 28 generative models released between 2022 and 2025, IncreFA achieves state-of-the-art attribution accuracy and 98.9% unseen detection under a temporally ordered open-set protocol.