Achieving Robustness in the Wild via Adversarial Mixing with Disentangled Representations
| Autor: | Timothy A. Mann, Po-Sen Huang, Krishnamurthy Dvijotham, Chongli Qin, Taylan Cemgil, Pushmeet Kohli, Sven Gowal |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2019 |
| Předmět: |
FOS: Computer and information sciences
Computer Science - Machine Learning Computer science Semantics (computer science) Generalization Computer Vision and Pattern Recognition (cs.CV) Computer Science - Computer Vision and Pattern Recognition Word error rate Machine Learning (stat.ML) 02 engineering and technology 010501 environmental sciences 01 natural sciences Machine Learning (cs.LG) Robustness (computer science) Statistics - Machine Learning 0202 electrical engineering electronic engineering information engineering Invariant (mathematics) 0105 earth and related environmental sciences business.industry Deep learning 020206 networking & telecommunications Invariant (physics) Bounded function Artificial intelligence business Algorithm |
| Zdroj: | CVPR |
| Popis: | Recent research has made the surprising finding that state-of-the-art deep learning models sometimes fail to generalize to small variations of the input. Adversarial training has been shown to be an effective approach to overcome this problem. However, its application has been limited to enforcing invariance to analytically defined transformations like $\ell_p$-norm bounded perturbations. Such perturbations do not necessarily cover plausible real-world variations that preserve the semantics of the input (such as a change in lighting conditions). In this paper, we propose a novel approach to express and formalize robustness to these kinds of real-world transformations of the input. The two key ideas underlying our formulation are (1) leveraging disentangled representations of the input to define different factors of variations, and (2) generating new input images by adversarially composing the representations of different images. We use a StyleGAN model to demonstrate the efficacy of this framework. Specifically, we leverage the disentangled latent representations computed by a StyleGAN model to generate perturbations of an image that are similar to real-world variations (like adding make-up, or changing the skin-tone of a person) and train models to be invariant to these perturbations. Extensive experiments show that our method improves generalization and reduces the effect of spurious correlations (reducing the error rate of a "smile" detector by 21% for example). Accepted at CVPR 2020 |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|