An ensemble of three pretrained vision transformers trained with supervised contrastive loss and majority voting reports 95.83% validation accuracy on the DFWild-Cup 2025 deepfake detection dataset.
Eva-02: A visual representation for neon genesis,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
DFCon: Attention-Driven Supervised Contrastive Learning for Robust Deepfake Detection
An ensemble of three pretrained vision transformers trained with supervised contrastive loss and majority voting reports 95.83% validation accuracy on the DFWild-Cup 2025 deepfake detection dataset.