AdvProp is an adversarial training scheme which treats adversarial examples as additional examples, to prevent overfitting. Key to the method is the usage of a separate auxiliary batch norm for adversarial examples, as they have different underlying distributions to normal examples.
Source: Adversarial Examples Improve Image RecognitionPaper | Code | Results | Date | Stars |
---|
Task | Papers | Share |
---|---|---|
Domain Generalization | 2 | 18.18% |
Image Classification | 2 | 18.18% |
Feature Importance | 1 | 9.09% |
Object Detection | 1 | 9.09% |
Adversarial Attack | 1 | 9.09% |
BIG-bench Machine Learning | 1 | 9.09% |
Drug Discovery | 1 | 9.09% |
Molecular Property Prediction | 1 | 9.09% |
Property Prediction | 1 | 9.09% |