Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

FOCA: Frequency-Oriented Cross-Domain Forgery Detection, Localization and Explanation via Multi-Modal Large Language Model

Zhou Liu, Tonghua Su, Hongshi Zhang, Fuxiang Yang, Donglin Di, Yang Song, Lei Fan

TL;DR

FOCA, a multimodal large language model-based framework that integrates discriminative features from both the RGB spatial and frequency domains via a cross-attention fusion module, is proposed, which enables accurate forgery detection and localization while providing explicit, human-interpretable cross-domain explanations.

Abstract

Advances in image tampering techniques, particularly generative models, pose significant challenges to media verification, digital forensics, and public trust. Existing image forgery detection and localization (IFDL) methods suffer from two key limitations: over-reliance on semantic content while neglecting textural cues, and limited interpretability of subtle low-level tampering traces. To address these issues, we propose FOCA, a multimodal large language model-based framework that integrates discriminative features from both the RGB spatial and frequency domains via a cross-attention fusion module. This design enables accurate forgery detection and localization while providing explicit, human-interpretable cross-domain explanations. We further introduce FSE-Set, a large-scale dataset with diverse authentic and tampered images, pixel-level masks, and dual-domain annotations. Extensive experiments show that FOCA outperforms state-of-the-art methods in detection performance and interpretability across both spatial and frequency domains.

FOCA: Frequency-Oriented Cross-Domain Forgery Detection, Localization and Explanation via Multi-Modal Large Language Model

TL;DR

FOCA, a multimodal large language model-based framework that integrates discriminative features from both the RGB spatial and frequency domains via a cross-attention fusion module, is proposed, which enables accurate forgery detection and localization while providing explicit, human-interpretable cross-domain explanations.

Abstract

Advances in image tampering techniques, particularly generative models, pose significant challenges to media verification, digital forensics, and public trust. Existing image forgery detection and localization (IFDL) methods suffer from two key limitations: over-reliance on semantic content while neglecting textural cues, and limited interpretability of subtle low-level tampering traces. To address these issues, we propose FOCA, a multimodal large language model-based framework that integrates discriminative features from both the RGB spatial and frequency domains via a cross-attention fusion module. This design enables accurate forgery detection and localization while providing explicit, human-interpretable cross-domain explanations. We further introduce FSE-Set, a large-scale dataset with diverse authentic and tampered images, pixel-level masks, and dual-domain annotations. Extensive experiments show that FOCA outperforms state-of-the-art methods in detection performance and interpretability across both spatial and frequency domains.
Paper Structure (11 sections, 8 equations, 3 figures, 2 tables)

This paper contains 11 sections, 8 equations, 3 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Method
  3. Overview
  4. Frequency Attention Fusion
  5. FSE-Set Dataset
  6. Training Objectives
  7. Experiments
  8. Experimental built upon
  9. Results and Discussion
  10. Conclusion
  11. Acknowledgents

Figures (3)

  • Figure 1: Overview of FOCA. It receives an image $x_{img}$ and a text instruction $x_{txt}$ and produces a textual response. FOCA uses the [CLS] token for tamper classification and the [SEG] token for pixel-level mask generation.
  • Figure 2: AI-Generated manipulations pipeline.
  • Figure 3: Comparative results of the pre-trained MLLMs and FOCA in tampering explanation capabilities on the FSE-Set.