A black-box, data-free method detects and removes patch-based trojan triggers from test images for self-supervised learning encoders.
The data augmentation operations used to process the pre-training data include RandomResizedCrop, RandomHorizontalFlip, ColorJitter, and RandomGrayScale
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CR 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
TrojanDec: Data-free Detection of Trojan Inputs in Self-supervised Learning
A black-box, data-free method detects and removes patch-based trojan triggers from test images for self-supervised learning encoders.