A CNN based detector and segmenter trained using compression rate variations of pristine vs spliced images and blue channel of the ELAs vs Ground Truth masks respectively.
Datasets used : CASIA1 | CASIA1GroundTruth
- Install Python 3.8
- Run
pip install --upgrade pip - Goto GUI folder
- Run
pip install -r requirements.txt - Start the program with
python FakeImageDetector.py
For Django Web App
- Run
pip install -r requirements.txt - Start the server with
python manage.py runserver - Go to
localhost:8000in your browser.