MGFFD-VLM combines quality-aware LoRA experts, multi-granularity prompts, and a three-stage training plan to improve explainable deepfake detection on the extended DD-VQA+ dataset.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
MGFFD-VLM: Multi-Granularity Prompt Learning for Face Forgery Detection with VLM
MGFFD-VLM combines quality-aware LoRA experts, multi-granularity prompts, and a three-stage training plan to improve explainable deepfake detection on the extended DD-VQA+ dataset.