Revealing the Distributional Vulnerability of Discriminators by Implicit Generators

• Published in: IEEE transactions on pattern analysis and machine intelligence Vol. 45; no. 7; pp. 8888 - 8901
• Main Authors: Zhao, Zhilin, Cao, Longbing, Lin, Kun-Yu
• Format: Journal Article
• Language: English
• Published: United States IEEE 01.07.2023; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Deep learning; Detectors; Discriminators; Entropy; Entropy (Information theory); generator-discriminator; Generators; Information theory; neural network vulnerability; Neural networks; out-of-distribution detection; Sensitivity; shannon entropy; Training
• ISSN: 0162-8828; 1939-3539; 2160-9292
• DOI: 10.1109/TPAMI.2022.3229318

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 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.