VAAS: Vision-Attention Anomaly Scoring for Image Manipulation Detection in Digital Forensics

TL;DR

VAAS introduces a dual-module framework combining Vision Transformers and patch-level analysis to detect and quantify image manipulations, improving explainability and reliability in digital forensics.

Contribution

It presents a novel hybrid anomaly scoring method that provides continuous, interpretable scores for image forgery detection, addressing limitations of existing approaches.

Findings

Achieves competitive F1 and IoU scores on benchmark datasets

Provides attention-guided anomaly maps for visual explainability

Enhances quantification of manipulation severity

Abstract

Recent advances in AI-driven image generation have introduced new challenges for verifying the authenticity of digital evidence in forensic investigations. Modern generative models can produce visually consistent forgeries that evade traditional detectors based on pixel or compression artefacts. Most existing approaches also lack an explicit measure of anomaly intensity, which limits their ability to quantify the severity of manipulation. This paper introduces Vision-Attention Anomaly Scoring (VAAS), a novel dual-module framework that integrates global attention-based anomaly estimation using Vision Transformers (ViT) with patch-level self-consistency scoring derived from SegFormer embeddings. The hybrid formulation provides a continuous and interpretable anomaly score that reflects both the location and degree of manipulation. Evaluations on the DF2023 and CASIA v2.0 datasets demonstrate that VAAS achieves competitive F1 and IoU performance, while enhancing visual explainability through attention-guided anomaly maps. The framework bridges quantitative detection with human-understandable reasoning, supporting transparent and reliable image integrity assessment. The source code for all experiments and corresponding materials for reproducing the results are available open source.