Decoder-free Robustness Disentanglement without (Additional) Supervision
arXiv:2007.01356
Abstract
Adversarial Training (AT) is proposed to alleviate the adversarial vulnerability of machine learning models by extracting only robust features from the input, which, however, inevitably leads to severe accuracy reduction as it discards the non-robust yet useful features. This motivates us to preserve both robust and non-robust features and separate them with disentangled representation learning. Our proposed Adversarial Asymmetric Training (AAT) algorithm can reliably disentangle robust and non-robust representations without additional supervision on robustness. Empirical results show our method does not only successfully preserve accuracy by combining two representations, but also achieve much better disentanglement than previous work.
References in corpus (19)
- Auto-Encoding Variational Bayes
- Intriguing properties of neural networks
- Wide Residual Networks
- Theoretically Principled Trade-off between Robustness and Accuracy
- Challenging Common Assumptions in the Unsupervised Learning of Disentangled Representations
- Deep Convolutional Inverse Graphics Network
- Disentangling by Factorising
- Adversarial Attacks and Defences: A Survey
- Adversarial Examples Are Not Bugs, They Are Features
- On the (Statistical) Detection of Adversarial Examples
- Robustness May Be at Odds with Accuracy
- Disentangling factors of variation in deep representations using adversarial training
- On Detecting Adversarial Perturbations
- Are Labels Required for Improving Adversarial Robustness?
- You Only Propagate Once: Accelerating Adversarial Training via Maximal Principle
- Adversarial Robustness as a Prior for Learned Representations
- Adversarial Examples Improve Image Recognition
- Feature Squeezing Mitigates and Detects Carlini/Wagner Adversarial Examples
- A Causal View on Robustness of Neural Networks