A model-agnostic input preprocessing pipeline of random resizing, median filtering, and JPEG compression limits adversarial efficiency-degradation attacks on token-pruning Vision Transformers to within 3.4% of unattacked GFLOPs.
Slowformer: Adversarial attack on compute and energy consumption of efficient vision transformers,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CR 1years
2026 1verdicts
CONDITIONAL 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
MOAT: Model-Agnostic Randomized Transformations for preventing Efficiency Degradation Attacks on ViTs
A model-agnostic input preprocessing pipeline of random resizing, median filtering, and JPEG compression limits adversarial efficiency-degradation attacks on token-pruning Vision Transformers to within 3.4% of unattacked GFLOPs.