FakeShield: Explainable Image Forgery Detection and Localization via Multi-modal Large Language Models
Zhipei Xu, Xuanyu Zhang, Runyi Li, Zecheng Tang, Qing Huang, Jian Zhang
School of Electronic and Computer Engineering, Peking University
💡 We also have other Copyright Protection projects that may interest you ✨.
AvatarShield: Visual Reinforcement Learning for Human-Centric Video Forgery Detection
Zhipei Xu, Xuanyu Zhang, Xing Zhou, Jian Zhang
EditGuard: Versatile Image Watermarking for Tamper Localization and Copyright Protection [CVPR 2024]
Xuanyu Zhang, Runyi Li, Jiwen Yu, Youmin Xu, Weiqi Li, Jian Zhang
OmniGuard: Hybrid Manipulation Localization via Augmented Versatile Deep Image Watermarking [CVPR 2025]
Xuanyu Zhang, Zecheng Tang, Zhipei Xu, Runyi Li, Youmin Xu, Bin Chen, Feng Gao, Jian Zhang
- [2026.02.21] 🔥🔥🔥 We have updated the SD_Inpaint dataset on Hugging Face, and you can access it from here.
- [2025.04.23] 🤗 We have open-sourced the MMTD-Set-34k dataset on Hugging Face, and you can access it from here.
- [2025.02.14] 🤗 We
are progressively open-sourcinghave open-sourced all code & pre-trained model weights. Welcome to watch 👀 this repository for the latest updates. - [2025.01.23] 🎉🎉🎉 Our FakeShield has been accepted at ICLR 2025!
- [2024.10.03] 🔥 We have released FakeShield: Explainable Image Forgery Detection and Localization via Multi-modal Large Language Models. We present explainable IFDL tasks, constructing the MMTD-Set dataset and the FakeShield framework. Check out the paper. The code and dataset are coming soon
FakeShield is a novel multi-modal framework designed for explainable image forgery detection and localization (IFDL). Unlike traditional black-box IFDL methods, FakeShield integrates multi-modal large language models (MLLMs) to analyze manipulated images, generate tampered region masks, and provide human-understandable explanations based on pixel-level artifacts and semantic inconsistencies. To improve generalization across diverse forgery types, FakeShield introduces domain tags, which guide the model to recognize different manipulation techniques effectively. Additionally, we construct MMTD-Set, a richly annotated dataset containing multi-modal descriptions of manipulated images, fostering better interpretability. Through extensive experiments, FakeShield demonstrates superior performance in detecting and localizing various forgeries, including copy-move, splicing, removal, DeepFake, and AI-generated manipulations.
-
FakeShield Introduction. We introduce FakeShield, a multi-modal framework for explainable image forgery detection and localization, which is the first to leverage MLLMs for the IFDL task. We also propose Domain Tag-guided Explainable Forgery Detection Module(DTE-FDM) and Multimodal Forgery Localization Module (MFLM) to improve the generalization and robustness of the models
-
Novel Explainable-IFDL Task. We propose the first explainable image forgery detection and localization (e-IFDL) task, addressing the opacity of traditional IFDL methods by providing both pixel-level and semantic-level explanations.
-
MMTD-Set Dataset Construction. We create the MMTD-Set by enriching existing IFDL datasets using GPT-4o, generating high-quality “image-mask-description” triplets for enhanced multimodal learning.
Note: If you want to reproduce the results from our paper, please prioritize using the Docker image to set up the environment. For more details, see this issue.
-
Ensure your environment meets the following requirements:
- Python == 3.9
- Pytorch == 1.13.0
- CUDA Version == 11.6
-
Clone the repository:
git clone https://github.com/zhipeixu/FakeShield.git cd FakeShield -
Install dependencies:
apt update && apt install git pip install -r requirements.txt ## Install MMCV git clone https://github.com/open-mmlab/mmcv cd mmcv git checkout v1.4.7 MMCV_WITH_OPS=1 pip install -e .
-
Install DTE-FDM:
cd ../DTE-FDM pip install -e . pip install -e ".[train]" pip install flash-attn --no-build-isolation
-
Pull the pre-built Docker image:
docker pull zhipeixu/mflm:v1.0 docker pull zhipeixu/dte-fdm:v1.0
-
Clone the repository:
git clone https://github.com/zhipeixu/FakeShield.git cd FakeShield -
Run the container:
docker run --gpus all -it --rm \ -v $(pwd):/workspace/FakeShield \ zhipeixu/dte-fdm:latest /bin/bash docker run --gpus all -it --rm \ -v $(pwd):/workspace/FakeShield \ zhipeixu/mflm:latest /bin/bash -
Inside the container, navigate to the repository:
cd /workspace/FakeShield -
Install MMCV:
git clone https://github.com/open-mmlab/mmcv
-
Download FakeShield weights from Hugging Face
The model weights consist of three parts:
DTE-FDM,MFLM, andDTG. For convenience, we have packaged them together and uploaded them to the Hugging Face repository.We recommend using
huggingface_hubto download the weights:pip install huggingface_hub huggingface-cli download --resume-download zhipeixu/fakeshield-v1-22b --local-dir weight/
-
Download pretrained SAM weight
In MFLM, we will use the SAM pre-training weights. You can use
wgetto download thesam_vit_h_4b8939.pthmodel:wget https://huggingface.co/ybelkada/segment-anything/resolve/main/checkpoints/sam_vit_h_4b8939.pth -P weight/
-
Ensure the weights are placed correctly
Organize your
weight/folder as follows:FakeShield/ ├── weight/ │ ├── fakeshield-v1-22b/ │ │ ├── DTE-FDM/ │ │ ├── MFLM/ │ │ ├── DTG.pth │ ├── sam_vit_h_4b8939.pth
You can quickly run the demo script by executing:
bash scripts/cli_demo.shThe cli_demo.sh script allows customization through the following environment variables:
WEIGHT_PATH: Path to the FakeShield weight directory (default:./weight/fakeshield-v1-22b)IMAGE_PATH: Path to the input image (default:./playground/image/Sp_D_CRN_A_ani0043_ani0041_0373.jpg)DTE_FDM_OUTPUT: Path for saving the DTE-FDM output (default:./playground/DTE-FDM_output.jsonl)MFLM_OUTPUT: Path for saving the MFLM output (default:./playground/DTE-FDM_output.jsonl)
Modify these variables to suit different use cases.
The training dataset consists of three types of data:
- PhotoShop Manipulation Dataset: CASIAv2, Fantastic Reality
- DeepFake Manipulation Dataset: FFHQ, FaceAPP
- AIGC-Editing Manipulation Dataset: SD_inpaint Dataset
- MMTD-Set Dataset: MMTD-Set
The validation dataset consists of three types of data:
- PhotoShop Manipulation Dataset: CASIA1+, IMD2020, Columbia, coverage, NIST16, DSO, Korus
- DeepFake Manipulation Dataset: FFHQ, FaceAPP
- AIGC-Editing Manipulation Dataset: SD_inpaint Dataset
- MMTD-Set Dataset: MMTD-Set
Download them from the above links and organize them as follows:
dataset/
├── photoshop/ # PhotoShop Manipulation Dataset
│ ├── CASIAv2_Tp/ # CASIAv2 Tampered Images
│ │ ├── image/
│ │ └── mask/
│ ├── CASIAv2_Au/ # CASIAv2 Authentic Images
│ │ └── image/
│ ├── FR_Tp/ # Fantastic Reality Tampered Images
│ │ ├── image/
│ │ └── mask/
│ ├── FR_Au/ # Fantastic Reality Authentic Images
│ │ └── image/
│ ├── CASIAv1+_Tp/ # CASIAv1+ Tampered Images
│ │ ├── image/
│ │ └── mask/
│ ├── CASIAv1+_Au/ # CASIAv1+ Authentic Images
│ │ └── image/
│ ├── IMD2020_Tp/ # IMD2020 Tampered Images
│ │ ├── image/
│ │ └── mask/
│ ├── IMD2020_Au/ # IMD2020 Authentic Images
│ │ └── image/
│ ├── Columbia/ # Columbia Dataset
│ │ ├── image/
│ │ └── mask/
│ ├── coverage/ # Coverage Dataset
│ │ ├── image/
│ │ └── mask/
│ ├── NIST16/ # NIST16 Dataset
│ │ ├── image/
│ │ └── mask/
│ ├── DSO/ # DSO Dataset
│ │ ├── image/
│ │ └── mask/
│ └── Korus/ # Korus Dataset
│ ├── image/
│ └── mask/
│
├── deepfake/ # DeepFake Manipulation Dataset
│ ├── FaceAPP_Train/ # FaceAPP Training Data
│ │ ├── image/
│ │ └── mask/
│ ├── FaceAPP_Val/ # FaceAPP Validation Data
│ │ ├── image/
│ │ └── mask/
│ ├── FFHQ_Train/ # FFHQ Training Data
│ │ └── image/
│ └── FFHQ_Val/ # FFHQ Validation Data
│ └── image/
│
├── aigc/ # AIGC Editing Manipulation Dataset
│ ├── SD_inpaint_Train/ # Stable Diffusion Inpainting Training Data
│ │ ├── image/
│ │ └── mask/
│ ├── SD_inpaint_Val/ # Stable Diffusion Inpainting Validation Data
│ │ ├── image/
│ │ └── mask/
│ ├── COCO2017_Train/ # COCO2017 Training Data
│ │ └── image/
│ └── COCO2017_Val/ # COCO2017 Validation Data
│ └── image/
│
└── MMTD_Set/ # Multi-Modal Tamper Description Dataset
└── MMTD-Set-34k.json # JSON Training FileYou can fine-tune DTE-FDM using LoRA with the following script:
bash ./scripts/DTE-FDM/finetune_lora.shThe script allows customization through the following environment variables:
OUTPUT_DIR: Directory for saving training outputDATA_PATH: Path to the training dataset (JSON format)WEIGHT_PATH: Path to the pre-trained weights
Modify these variables as needed to adapt the training process to different datasets and setups.
You can fine-tune MFLM using LoRA with the following script:
bash ./scripts/MFLM/finetune_lora.shThe script allows customization through the following environment variables:
OUTPUT_DIR: Directory for saving training outputDATA_PATH: Path to the training datasetWEIGHT_PATH: Path to the pre-trained weightsTRAIN_DATA_CHOICE: Selecting the training datasetVAL_DATA_CHOICE: Selecting the validation dataset
Modify these variables as needed to adapt the training process to different datasets and setups.
You can test FakeShield using the following script:
bash ./scripts/test.shThe script allows customization through the following environment variables:
WEIGHT_PATH: Path to the directory containing the FakeShield model weights.QUESTION_PATH: Path to the test dataset in JSONL format. This file can be generated using./playground/eval_jsonl.py.DTE_FDM_OUTPUT: Path for saving the output of the DTE-FDM model.MFLM_OUTPUT: Path for saving the output of the MFLM model.
Modify these variables as needed to adapt the evaluation process to different datasets and setups.
@inproceedings{xu2024fakeshield,
title={FakeShield: Explainable Image Forgery Detection and Localization via Multi-modal Large Language Models},
author={Xu, Zhipei and Zhang, Xuanyu and Li, Runyi and Tang, Zecheng and Huang, Qing and Zhang, Jian},
booktitle={International Conference on Learning Representations},
year={2025}
}We are thankful to LLaVA, groundingLMM, and LISA for releasing their models and code as open-source contributions.
