A dual-encoder deepfake detector pairs a frozen specialist with a LoRA-tuned MLLM, trained first via binary alignment then via RL to reward explain-then-classify behavior, yielding improved cross-dataset performance and interpretability.
arXiv preprint arXiv:2509.19841 , year=
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.CV 2years
2026 2representative citing papers
citing papers explorer
-
The Regularizing Power of Language-Training Deepfake Detectors
A dual-encoder deepfake detector pairs a frozen specialist with a LoRA-tuned MLLM, trained first via binary alignment then via RL to reward explain-then-classify behavior, yielding improved cross-dataset performance and interpretability.
- Venus-DeFakerOne: Unified Fake Image Detection & Localization