A unified framework for attributing model components, data, and training dynamics to model behavior.
Official code for CapST (ACM TOMM 2025): A Capsule Network and Temporal Attention-based framework for Deepfake Model Attribution. Achieves efficient and accurate source identification across DFDM and GAN-generated datasets.
This is the public version of my master's research work.
Official code for FAME: A lightweight spatio-temporal framework for Deepfake Model Attribution using VGG-based feature extraction and temporal attention. Outperforms prior methods across DFDM, FF++, and FakeAVCeleb datasets.
Add a description, image, and links to the model-attribution topic page so that developers can more easily learn about it.
To associate your repository with the model-attribution topic, visit your repo's landing page and select "manage topics."