Task
Video Classification
D
jai1th/dvf_trained_transferred_aegis
Video Classification·jai1th· 45
R(2+1)D + mixed-domain supervised-contrastive model for end-to-end AI-generated video detection. A DVF-trained R(2+1)D backbone is transferred to the AEGIS domain while retaining performance on DVF and GenVideo, then evaluated with a projection-space prototype protocol.
# pull & run locally
pip install mlforge-sdk && mlforge pull jai1th/dvf_trained_transferred_aegis
pip install mlforge-sdk && mlforge pull jai1th/dvf_trained_transferred_aegis
Model details
Provider
jai1th
Framework
pytorch
Size
661 MB
License
cc-by-nc-4.0
Downloads
45
Paper
arXiv:2410.23623
Updated
2026-06-16
About jai1th/dvf_trained_transferred_aegis
R(2+1)D + mixed-domain supervised-contrastive model for end-to-end AI-generated video detection. A DVF-trained R(2+1)D backbone is transferred to the AEGIS domain while retaining performance on DVF and GenVideo, then evaluated with a projection-space prototype protocol.