Robust Audio-Visual Deepfake Detection with Cross-Dataset Generalization Under Real-World Distortions

A robust audio-visual deepfake detector that generalises across datasets under real-world distortions.

A robust audio-visual deepfake detector that generalises across datasets under real-world distortions. This reference project walks through the whole job, from the problem and the design to the build and the evaluation, so that a developer can study how it works, run it and extend it.

It suits learners who want to work with images or video: building a dataset, training or fine-tuning a vision model, measuring its performance and wrapping it in a usable tool.

The main parts are dataset preparation and annotation, model training or fine-tuning, evaluation with standard metrics, plus visual examples, and an interface that works on uploaded images or live video. The reference build uses Python, OpenCV, PyTorch, and NumPy, with a clear project structure, setup instructions and sample data so that it can be run and extended.