Safer Classification by Synthesis

2017/11/22 by William Yang Wang, William Wang, Angelina Wang +8 · 1 voice · 1 citation
Computer Science · Mathematics · #Adversarial Robustness in Machine Learning #Anomaly Detection Techniques and Applications #Artificial Intelligence (cs.AI) #FOS: Computer and information sciences #Generative Adversarial Networks and Image Synthesis #Machine Learning (cs.LG) #Machine Learning (stat.ML) #cs.AI #cs.LG #stat.ML

paper · pdf · doi:10.48550/arxiv.1711.08534

openalex publication_date 2017/11/22 · arxiv published 2017/11/22 · arxiv created 2018/07/23 · arxiv updated 2018/07/25 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

The discriminative approach to classification using deep neural networks has become the de-facto standard in various fields. Complementing recent reservations about safety against adversarial examples, we show that conventional discriminative methods can easily be fooled to provide incorrect labels with very high confidence to out of distribution examples. We posit that a generative approach is the natural remedy for this problem, and propose a method for classification using generative models. At training time, we learn a generative model for each class, while at test time, given an example to classify, we query each generator for its most similar generation, and select the class corresponding to the most similar one. Our approach is general and can be used with expressive models such as GANs and VAEs. At test time, our method accurately "knows when it does not know," and provides resilience to out of distribution examples while maintaining competitive performance for standard examples.

Citations

• Auto-Encoding Variational Bayes
• Human-level concept learning through probabilistic program induction
• Weight Normalization: A Simple Reparameterization to Accelerate Training of Deep Neural Networks
• Adversarial Feature Learning
• Improved Techniques for Training GANs
• Towards Principled Methods for Training Generative Adversarial Networks
• BEGAN: Boundary Equilibrium Generative Adversarial Networks
• Distance-based Confidence Score for Neural Network Classifiers
• Novelty Detection with GAN
• Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning

Cited by

• Deep Variational Semi-Supervised Novelty Detection

Discussions

• Safer Classification by Synthesis [pdf] [hn, 2 points, 0 comments]

Related

• Are Generative Classifiers More Robust to Adversarial Attacks?
• Defense-GAN: Protecting Classifiers Against Adversarial Attacks Using Generative Models
• Versatile Auxiliary Classifier with Generative Adversarial Network (VAC+GAN)
• Understanding Classifier Mistakes with Generative Models
• Reject Illegal Inputs with Generative Classifier Derived from Any Discriminative Classifier