DL-Based Solution for Detection of Face-Swap Deep Fake Photos

https://github.com/shashwat39/deepfake-detection

Developed and optimized deep learning models (MesoNet, Swin Transformer, CNNs) to detect face-swap deep fake images, achieving high accuracy across multiple publicly available datasets.

Integrated MLOps pipeline with tools like Weights & Biases for experiment tracking, Hydra for configuration management, and DVC for dataset versioning to ensure smooth model development and deployment.

Implemented containerization with Docker and CI/CD workflows using GitHub Actions, enabling scalable model deployment and automated testing for reliable real-time predictions in production.