CapsFake: A Multimodal Capsule Network for Detecting Instruction-Guided Deepfakes
This work addresses the threat to digital image integrity from advanced deepfake technology, offering a robust detection framework for security applications, though it appears incremental as it builds on existing capsule network and multimodal approaches.
The paper tackles the problem of detecting instruction-guided deepfake image edits, which are subtle and context-aware manipulations, by proposing CapsFake, a multimodal capsule network that integrates visual, textual, and frequency-domain features. It achieves up to 20% higher detection accuracy than state-of-the-art methods, with rates above 94% under natural perturbations and 96% against adversarial attacks.
The rapid evolution of deepfake technology, particularly in instruction-guided image editing, threatens the integrity of digital images by enabling subtle, context-aware manipulations. Generated conditionally from real images and textual prompts, these edits are often imperceptible to both humans and existing detection systems, revealing significant limitations in current defenses. We propose a novel multimodal capsule network, CapsFake, designed to detect such deepfake image edits by integrating low-level capsules from visual, textual, and frequency-domain modalities. High-level capsules, predicted through a competitive routing mechanism, dynamically aggregate local features to identify manipulated regions with precision. Evaluated on diverse datasets, including MagicBrush, Unsplash Edits, Open Images Edits, and Multi-turn Edits, CapsFake outperforms state-of-the-art methods by up to 20% in detection accuracy. Ablation studies validate its robustness, achieving detection rates above 94% under natural perturbations and 96% against adversarial attacks, with excellent generalization to unseen editing scenarios. This approach establishes a powerful framework for countering sophisticated image manipulations.