FROB: Few-shot ROBust Model for Classification and Out-of-Distribution Detection

Abstract

Nowadays, classification and Out-of-Distribution (OoD) detection in the few-shot setting remain challenging aims due to rarity and the limited samples in the few-shot setting, and because of adversarial attacks. Accomplishing these aims is important for critical systems in safety, security, and defence. In parallel, OoD detection is challenging since deep neural network classifiers set high confidence to OoD samples away from the training data. To address such limitations, we propose the Few-shot ROBust (FROB) model for classification and few-shot OoD detection. We devise FROB for improved robustness and reliable confidence prediction for few-shot OoD detection. We generate the support boundary of the normal class distribution and combine it with few-shot Outlier Exposure (OE). We propose a self-supervised learning few-shot confidence boundary methodology based on generative and discriminative models. The contribution of FROB is the combination of the generated boundary in a self-supervised learning manner and the imposition of low confidence at this learned boundary. FROB implicitly generates strong adversarial samples on the boundary and forces samples from OoD, including our boundary, to be less confident by the classifier. FROB achieves generalization to unseen OoD with applicability to unknown, in the wild, test sets that do not correlate to the training datasets. To improve robustness, FROB redesigns OE to work even for zero-shots. By including our boundary, FROB reduces the threshold linked to the model's few-shot robustness; it maintains the OoD performance approximately independent of the number of few-shots. The few-shot robustness analysis evaluation of FROB on different sets and on One-Class Classification (OCC) data shows that FROB achieves competitive performance and outperforms benchmarks in terms of robustness to the outlier few-shot sample population and variability.

arXiv commentPaper, 22 pages, Figures, Tables

References in corpus (27)

• Improved Techniques for Training GANs
• A Simple Unified Framework for Detecting Out-of-Distribution Samples and Adversarial Attacks
• AutoAugment: Learning Augmentation Policies from Data
• Reliable evaluation of adversarial robustness with an ensemble of diverse parameter-free attacks
• Evidential Deep Learning to Quantify Classification Uncertainty
• Deep Anomaly Detection with Outlier Exposure
• Uncertainty Estimation Using a Single Deep Deterministic Neural Network
• CSI: Novelty Detection via Contrastive Learning on Distributionally Shifted Instances
• The Robust Manifold Defense: Adversarial Training using Generative Models
• Simple and Principled Uncertainty Estimation with Deterministic Deep Learning via Distance Awareness
• A Simple Fix to Mahalanobis Distance for Improving Near-OOD Detection
• Confidence-Aware Learning for Deep Neural Networks
• Poisoning Attacks with Generative Adversarial Nets
• Out-of-distribution Detection in Classifiers via Generation
• SSD: A Unified Framework for Self-Supervised Outlier Detection
• Safer Classification by Synthesis
• Quantifying Perceptual Distortion of Adversarial Examples
• Certifiably Adversarially Robust Detection of Out-of-Distribution Data
• Building robust classifiers through generation of confident out of distribution examples
• Tail of Distribution GAN (TailGAN): Generative-Adversarial-Network-Based Boundary Formation
• Boundary of Distribution Support Generator (BDSG): Sample Generation on the Boundary
• Are all outliers alike? On Understanding the Diversity of Outliers for Detecting OODs
• Analysis of Confident-Classifiers for Out-of-distribution Detection
• Detecting OODs as datapoints with High Uncertainty
• A General Framework For Detecting Anomalous Inputs to DNN Classifiers
• Provably Robust Detection of Out-of-distribution Data (almost) for free
• OMASGAN: Out-of-Distribution Minimum Anomaly Score GAN for Sample Generation on the Boundary