UPDATE (07/2024): 🆕 MAJOR ENHANCEMENT: Added comprehensive AI-generated image detection capabilities! The toolbox now includes Python-based AI detection algorithms, unified classification system (AI-generated/Edited/Original), and effectiveness assessment tools. See evaluation/README_ENHANCED.md for details.

UPDATE (01/2019): We have now added a new Noise-based algorithm in the MATLAB toolbox (NOI5). Make sure to properly cite the original paper if you use it (see the corresponding README file).

UPDATE (07/2018): We have now added our novel CAGI tampering detection algorithm in both Matlab and Java. It is based on JPEG block grid inconsistencies, but also features significant post-processing to improve result quality. Feel free to try it out! You will find citation information in the README file in the coresponding Matlab folder.

image-forensics

This is an integrated framework for image forensic analysis, now enhanced with AI-generated content detection capabilities. The framework includes:

• 🧠 AI Detection Module: New Python-based algorithms for detecting AI-generated images (GANs, diffusion models, etc.)
• 🔬 Traditional Tampering Detection: Existing MATLAB implementations of splicing detection algorithms
• 🎯 Unified Classification: Comprehensive system classifying images as AI-generated, edited, or original
• 📊 Effectiveness Assessment: Tools to evaluate algorithm performance and identify enhancement needs

Quick Start with Enhanced Features

# Install basic requirements
pip install numpy

# Run effectiveness assessment
cd evaluation
python3 assess_effectiveness.py

# Test AI detection on an image
python3 simple_ai_demo.py path/to/image.jpg

# For full functionality, install additional dependencies
pip install opencv-python pillow
python3 unified_classifier.py path/to/image.jpg --verbose

See evaluation/README_ENHANCED.md for comprehensive documentation of the new AI detection capabilities.

Original Framework

The original framework includes a Java webservice with seven splicing detection algorithm implementations, plus additional forensic tools (located in java_service) and a Matlab algorithm evaluation framework with implementations of a large number of splicing detection algorithms (located in matlab_toolbox). An online demo providing implementations of many of these algorithms can be found here: Online demo.

Citations

Please cite the following paper in your publications if you use the Java implementations:

@inproceedings{zamp16,
  author = "Markos Zampoglou and Symeon Papadopoulos and Yiannis Kompatsiaris and Ruben Bouwmeester and Jochen Spangenberg",
  booktitle = "Social Media In the NewsRoom, {#SMNews16@CWSM}, Tenth International AAAI Conference on Web and Social Media workshops",
  title = "Web and Social Media Image Forensics for News Professionals",
  year = "2016",
}

If you use the Matlab implementations, use the following citation:

@article{zampAcc,
  author = "Markos Zampoglou and Symeon Papadopoulos and Yiannis Kompatsiaris",
  title = "A Large-Scale Evaluation of Splicing Localization Algorithms for Web Images",
  journal = "Multimedia Tools and Applications",
  doi = "10.1007/s11042-016-3795-2"
  pages= "Accepted for publication",
}

In either case, you must also cite the original algorithm publication. The README file within each Matlab algorithm subfolder contains the corresponding citation.

Contact Markos Zampoglou [email protected] for any further information.