VLMs formulate differentiable rewards from task-specific rules to enable test-time online LoRA optimization of VGMs, delivering 16.7-point gains on symbolic and general video reasoning benchmarks over VLM-as-solver and Best-of-N baselines.
Denoising diffusion probabilistic models
5 Pith papers cite this work. Polarity classification is still indexing.
5
Pith papers citing it
verdicts
UNVERDICTED 5representative citing papers
GenAI-FDIA benchmarks physics-informed generative models for stealthy FDIA synthesis on IEEE testbeds, reports high evasion rates, and introduces an inference-time harmoniser plus warm-up schedules to fix projection displacement and covariance collapse.
DDPO-VC applies diffusion denoising policy optimization with dual-teacher rewards to improve speaker de-identification while preserving cognitive utility on dementia speech benchmarks.
DreamAudio generates audio clips that incorporate user-specified personalized audio events from reference samples while remaining aligned with text prompts.
FMSD improves cross-dataset generalization in deepfake detection by using gradient-based layer masking to select forgery-sensitive weights and SVD to split them into preserved semantic and multiple learnable artifact subspaces with orthogonality constraints.
citing papers explorer
-
VLMs are Good Teachers for Video Reasoning via Adaptive Test-Time Optimization
VLMs formulate differentiable rewards from task-specific rules to enable test-time online LoRA optimization of VGMs, delivering 16.7-point gains on symbolic and general video reasoning benchmarks over VLM-as-solver and Best-of-N baselines.
-
GenAI-FDIA: Physics-Informed Generative Models for False Data Injection Attacks
GenAI-FDIA benchmarks physics-informed generative models for stealthy FDIA synthesis on IEEE testbeds, reports high evasion rates, and introduces an inference-time harmoniser plus warm-up schedules to fix projection displacement and covariance collapse.
-
DDPO-VC: Speaker De-Identification via Diffusion Denoising Policy Optimization
DDPO-VC applies diffusion denoising policy optimization with dual-teacher rewards to improve speaker de-identification while preserving cognitive utility on dementia speech benchmarks.
-
DreamAudio: Customized Text-to-Audio Generation with Diffusion Models
DreamAudio generates audio clips that incorporate user-specified personalized audio events from reference samples while remaining aligned with text prompts.
-
Generalizable Deepfake Detection Based on Forgery-aware Layer Masking and Multi-artifact Subspace Decomposition
FMSD improves cross-dataset generalization in deepfake detection by using gradient-based layer masking to select forgery-sensitive weights and SVD to split them into preserved semantic and multiple learnable artifact subspaces with orthogonality constraints.