Explainable Deepfake Detection with RL Enhanced Self-Blended Images

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Jiang, Ning, Zeng, Dingheng, Liu, Yanhong, Yi, Haiyang, Yu, Shijie, Weng, Minghe, Shen, Haifeng, Li, Ying
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909997753958400
author Jiang, Ning
Zeng, Dingheng
Liu, Yanhong
Yi, Haiyang
Yu, Shijie
Weng, Minghe
Shen, Haifeng
Li, Ying
author_facet Jiang, Ning
Zeng, Dingheng
Liu, Yanhong
Yi, Haiyang
Yu, Shijie
Weng, Minghe
Shen, Haifeng
Li, Ying
contents Most prior deepfake detection methods lack explainable outputs. With the growing interest in multimodal large language models (MLLMs), researchers have started exploring their use in interpretable deepfake detection. However, a major obstacle in applying MLLMs to this task is the scarcity of high-quality datasets with detailed forgery attribution annotations, as textual annotation is both costly and challenging - particularly for high-fidelity forged images or videos. Moreover, multiple studies have shown that reinforcement learning (RL) can substantially enhance performance in visual tasks, especially in improving cross-domain generalization. To facilitate the adoption of mainstream MLLM frameworks in deepfake detection with reduced annotation cost, and to investigate the potential of RL in this context, we propose an automated Chain-of-Thought (CoT) data generation framework based on Self-Blended Images, along with an RL-enhanced deepfake detection framework. Extensive experiments validate the effectiveness of our CoT data construction pipeline, tailored reward mechanism, and feedback-driven synthetic data generation approach. Our method achieves performance competitive with state-of-the-art (SOTA) approaches across multiple cross-dataset benchmarks. Implementation details are available at https://github.com/deon1219/rlsbi.
format Preprint
id arxiv_https___arxiv_org_abs_2601_15624
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Explainable Deepfake Detection with RL Enhanced Self-Blended Images
Jiang, Ning
Zeng, Dingheng
Liu, Yanhong
Yi, Haiyang
Yu, Shijie
Weng, Minghe
Shen, Haifeng
Li, Ying
Computer Vision and Pattern Recognition
Most prior deepfake detection methods lack explainable outputs. With the growing interest in multimodal large language models (MLLMs), researchers have started exploring their use in interpretable deepfake detection. However, a major obstacle in applying MLLMs to this task is the scarcity of high-quality datasets with detailed forgery attribution annotations, as textual annotation is both costly and challenging - particularly for high-fidelity forged images or videos. Moreover, multiple studies have shown that reinforcement learning (RL) can substantially enhance performance in visual tasks, especially in improving cross-domain generalization. To facilitate the adoption of mainstream MLLM frameworks in deepfake detection with reduced annotation cost, and to investigate the potential of RL in this context, we propose an automated Chain-of-Thought (CoT) data generation framework based on Self-Blended Images, along with an RL-enhanced deepfake detection framework. Extensive experiments validate the effectiveness of our CoT data construction pipeline, tailored reward mechanism, and feedback-driven synthetic data generation approach. Our method achieves performance competitive with state-of-the-art (SOTA) approaches across multiple cross-dataset benchmarks. Implementation details are available at https://github.com/deon1219/rlsbi.
title Explainable Deepfake Detection with RL Enhanced Self-Blended Images
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2601.15624