Proto-LeakNet: Towards Signal-Leak Aware Attribution in Synthetic Human Face Imagery

TL;DR

Proto-LeakNet is a novel framework that detects and attributes synthetic human face images by leveraging signal-leaks in diffusion models, achieving high accuracy and robustness in identifying both known and unseen generators.

Contribution

It introduces a signal-leak-aware attribution method operating in the latent domain of diffusion models, with a novel interpretability and open-set evaluation capability.

Findings

Achieves a Macro AUC of 98.13% on attribution tasks.

Outperforms state-of-the-art methods in robustness and accuracy.

Effectively distinguishes between real, known, and unseen synthetic images.

Abstract

The growing sophistication of synthetic image and deepfake generation models has turned source attribution and authenticity verification into a critical challenge for modern computer vision systems. Recent studies suggest that diffusion pipelines unintentionally imprint persistent statistical traces, known as signal-leaks, within their outputs, particularly in latent representations. Building on this observation, we propose Proto-LeakNet, a signal-leak-aware and interpretable attribution framework that integrates closed-set classification with a density-based open-set evaluation on the learned embeddings, enabling analysis of unseen generators without retraining. Acting in the latent domain of diffusion models, our method re-simulates partial forward diffusion to expose residual generator-specific cues. A temporal attention encoder aggregates multi-step latent features, while a feature-weighted prototype head structures the embedding space and enables transparent attribution. Trained solely on closed data and achieving a Macro AUC of 98.13%, Proto-LeakNet learns a latent geometry that remains robust under post-processing, surpassing state-of-the-art methods, and achieves strong separability both between real images and known generators, and between known and unseen ones. The codebase will be available after acceptance.