SEED: A Large-Scale Benchmark for Provenance Tracing in Sequential Deepfake Facial Edits

2026-04-12

AI Summary

This work addresses the limitation of existing deepfake datasets, which predominantly support only single-step editing and thus hinder provenance analysis of multi-step manipulations. To bridge this gap, the authors introduce SEED, the first large-scale benchmark for sequential editing, comprising over 90,000 facial images edited through 1–4 steps using diffusion models, accompanied by fine-grained annotations including editing order, textual instructions, operation masks, and source generative models. Building upon this benchmark, they propose FAITH, a baseline model that integrates spatial and wavelet frequency-domain features, revealing for the first time the critical role of high-frequency signals in detecting multi-step editing traces. Experiments demonstrate that methods relying solely on spatial information are inherently limited, whereas FAITH leverages frequency-domain cues to significantly improve attribution accuracy and maintains robustness under image degradation.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Natural Language Processing: Fact-Checking / Misinformation Detection (NLP Focus)
• Machine Learning: Large Multimodal Models (LMMs)

Application Category

• Security and Privacy: Data transparency and provenance
• Web Mining and Content Analysis: Web data provenance, reliability, and authenticity
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems

Abstract

Deepfake content on social networks is increasingly produced through multiple \emph{sequential} edits to biometric data such as facial imagery. Consequently, the final appearance of an image often reflects a latent chain of operations rather than a single manipulation. Recovering these editing histories is essential for visual provenance analysis, misinformation auditing, and forensic or platform moderation workflows that must trace the origin and evolution of AI-generated media. However, existing datasets predominantly focus on single-step editing and overlook the cumulative artifacts introduced by realistic multi-step pipelines. To address this gap, we introduce Sequential Editing in Diffusion (\textbf{SEED}), a large-scale benchmark for sequential provenance tracing in facial imagery. SEED contains over 90K images constructed via one to four sequential attribute edits using diffusion-based editing pipelines, with fine-grained annotations including edit order, textual instructions, manipulation masks, and generation models. These metadata enable step-wise evidence analysis and support forgery detection, sequence prediction. To benchmark the challenges posed by SEED, we evaluate representative analysis strategies and observe that spatial-only approaches struggle under subtle and distributed diffusion artifacts, especially when such artifacts accumulate across multiple edits. Motivated by this observation, we further establish \textbf{FAITH}, a frequency-aware Transformer baseline that aggregates spatial and frequency-domain cues to identify and order latent editing events. Results show that high-frequency signals, particularly wavelet components, provide effective cues even under image degradation. Overall, SEED facilitates systematic study of sequential provenance tracing and evidence aggregation for trustworthy analysis of AI-generated visual content.

Problem

• provenance tracing
• sequential deepfake
• facial editing
• diffusion artifacts
• visual forensics

Innovation

• sequential deepfake editing
• provenance tracing
• diffusion-based editing
• frequency-aware Transformer
• visual forensics