ANL uses diffusion noise prediction and attention to regularize deepfake detectors for better generalization to unseen synthesis methods without added inference cost.
Weiss, Niru Maheswaranathan, and Surya Ganguli
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
representative citing papers
OFA-Diffusion Compression trains diffusion models once to yield multiple size-specific compressed subnetworks via restricted candidate spaces, importance-based channel allocation, and reweighting.
NeurBench is a benchmark suite that quantifies drift via a drift factor and generates data/workloads to evaluate learned database components under controllable drift scenarios.
citing papers explorer
-
Deepfake Detection Generalization with Diffusion Noise
ANL uses diffusion noise prediction and attention to regularize deepfake detectors for better generalization to unseen synthesis methods without added inference cost.
-
OFA-Diffusion Compression: Compressing Diffusion Model in One-Shot Manner
OFA-Diffusion Compression trains diffusion models once to yield multiple size-specific compressed subnetworks via restricted candidate spaces, importance-based channel allocation, and reweighting.
-
NeurBench: A Benchmark Suite for Learned Database Components with Drift Modeling
NeurBench is a benchmark suite that quantifies drift via a drift factor and generates data/workloads to evaluate learned database components under controllable drift scenarios.