DFA-CON trains a ResNet-50 with supervised contrastive loss to embed originals and their AI-forged versions close together, achieving the best reported F1 on the DeepfakeArt benchmark among the tested models.
Results indicate that using embeddings directly from the encoder output in R2048 yields the highest scores across all metrics
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
DFA-CON: A Contrastive Learning Approach for Detecting Copyright Infringement in DeepFake Art
DFA-CON trains a ResNet-50 with supervised contrastive loss to embed originals and their AI-forged versions close together, achieving the best reported F1 on the DeepfakeArt benchmark among the tested models.