Revealing the Distributional Vulnerability of Discriminators by Implicit Generators

TL;DR

This paper introduces FIG, a method that fine-tunes discriminators to better detect out-of-distribution samples by generating implicit OOD examples without retraining, enhancing robustness and safety in deep learning.

Contribution

The paper proposes a novel implicit generator-based fine-tuning approach for discriminators, improving OOD detection without additional training.

Findings

FIG achieves state-of-the-art OOD detection performance.

The method improves discriminator robustness against OOD samples.

Implicit generators effectively produce OOD samples for regularization.

Abstract

In deep neural learning, a discriminator trained on in-distribution (ID) samples may make high-confidence predictions on out-of-distribution (OOD) samples. This triggers a significant matter for robust, trustworthy and safe deep learning. The issue is primarily caused by the limited ID samples observable in training the discriminator when OOD samples are unavailable. We propose a general approach for \textit{fine-tuning discriminators by implicit generators} (FIG). FIG is grounded on information theory and applicable to standard discriminators without retraining. It improves the ability of a standard discriminator in distinguishing ID and OOD samples by generating and penalizing its specific OOD samples. According to the Shannon entropy, an energy-based implicit generator is inferred from a discriminator without extra training costs. Then, a Langevin dynamic sampler draws specific OOD samples for the implicit generator. Lastly, we design a regularizer fitting the design principle of the implicit generator to induce high entropy on those generated OOD samples. The experiments on different networks and datasets demonstrate that FIG achieves the state-of-the-art OOD detection performance.