Safer Classification by Synthesis
William Wang, Angelina Wang, Aviv Tamar, Xi Chen、Pieter Abbeel Dept. of Electrical Engineering and Computer Sciences UC Berkeley Berkeley, CA 94709
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.
中文速览
判别式深度学习分类器（discriminative classifier）在遇到训练数据之外的样本时，会以极高的置信度给出错误预测——即"不知道却自以为知道"——这一安全隐患在自动驾驶等高风险场景中尤为危险。为此，作者提出用生成模型（generative model）来做分类：训练阶段为每个类别分别训练一个生成器（GAN或VAE），测试阶段对输入图像在各类别的生成器潜空间中搜索最相似的生成结果，相似度最高的那个类别即为预测结果。实验表明，这种方法在保持正常样本分类精度的同时，对分布外（out-of-distribution）样本能够显著降低错误置信度，真正做到"不会的题不乱猜"。该工作将费曼"能创造才算真正理解"的理念引入分类任务，为构建更可信、更安全的AI系统提供了一条基于生成理解的新路径。
原文 arXiv:1711.08534;中英对照 + 大白话阅读 https://aha.fim.ai/paper/1711.08534v2