SSD: A Unified Framework for Self-Supervised Outlier Detection

2021/03/22 by Vikash Sehwag, Mung Chiang, Sehwag, Vikash +3 · 19 citations
Computer Science · #Adversarial Robustness in Machine Learning #Anomaly Detection Techniques and Applications #Artificial Intelligence (cs.AI) #Computer Vision and Pattern Recognition (cs.CV) #Domain Adaptation and Few-Shot Learning #FOS: Computer and information sciences #Machine Learning (cs.LG)

paper · pdf · doi:10.48550/arxiv.2103.12051

openalex publication_date 2021/03/22 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

We ask the following question: what training information is required to design an effective outlier/out-of-distribution (OOD) detector, i.e., detecting samples that lie far away from the training distribution? Since unlabeled data is easily accessible for many applications, the most compelling approach is to develop detectors based on only unlabeled in-distribution data. However, we observe that most existing detectors based on unlabeled data perform poorly, often equivalent to a random prediction. In contrast, existing state-of-the-art OOD detectors achieve impressive performance but require access to fine-grained data labels for supervised training. We propose SSD, an outlier detector based on only unlabeled in-distribution data. We use self-supervised representation learning followed by a Mahalanobis distance based detection in the feature space. We demonstrate that SSD outperforms most existing detectors based on unlabeled data by a large margin. Additionally, SSD even achieves performance on par, and sometimes even better, with supervised training based detectors. Finally, we expand our detection framework with two key extensions. First, we formulate few-shot OOD detection, in which the detector has access to only one to five samples from each class of the targeted OOD dataset. Second, we extend our framework to incorporate training data labels, if available. We find that our novel detection framework based on SSD displays enhanced performance with these extensions, and achieves state-of-the-art performance. Our code is publicly available at https://github.com/inspire-group/SSD.

Citations

• Auto-Encoding Variational Bayes
• Procedures for Detecting Outlying Observations in Samples
• Estimating the Support of a High-Dimensional Distribution
• A Simple Unified Framework for Detecting Out-of-Distribution Samples and Adversarial Attacks
• Reducing Network Agnostophobia
• Deep Anomaly Detection with Outlier Exposure
• PixelCNN++: Improving the PixelCNN with Discretized Logistic Mixture Likelihood and Other Modifications
• Classification-Based Anomaly Detection for General Data
• Momentum Contrast for Unsupervised Visual Representation Learning
• Learning Multiple Layers of Features from Tiny Images
• Deep Learning for Anomaly Detection: A Review

Cited by

• Predictive Sample Assignment for Semantically Coherent Out-of-Distribution Detection
• Rethinking Jailbreak Detection of Large Vision Language Models with Representational Contrastive Scoring
• Domain Feature Collapse: Implications for Out-of-Distribution Detection and Solutions
• Fewer Tokens, Greater Scaling: Self-Adaptive Visual Bases for Efficient and Expansive Representation Learning
• AutoSciDACT: Automated Scientific Discovery through Contrastive Embedding and Hypothesis Testing
• A Multi-dimensional Semantic Surprise Framework Based on Low-Entropy Semantic Manifolds for Fine-Grained Out-of-Distribution Detection
• Towards Robust Artificial Intelligence: Self-Supervised Learning Approach for Out-of-Distribution Detection
• Local Background Features Matter in Out-of-Distribution Detection
• Equipping Vision Foundation Model with Mixture of Experts for Out-of-Distribution Detection
• Medix: Out-of-Distribution Detection from Unlabeled Wild Data via Robust Gradient Statistics
• An Investigation into the Performance of Non-Contrastive Self-Supervised Learning Methods for Network Intrusion Detection
• Sharpness-Aware Geometric Defense for Robust Out-Of-Distribution Detection
• Draw a Portrait of Your Graph Data: An Instance-Level Profiling Framework for Graph-Structured Data
• Prompt Optimization Meets Subspace Representation Learning for Few-shot Out-of-Distribution Detection
• Tackling the Noisy Elephant in the Room: Label Noise-robust Out-of-Distribution Detection via Loss Correction and Low-rank Decomposition
• Frustratingly Easy Feature Reconstruction for Out-of-Distribution Detection
• Activation Subspaces for Out-of-Distribution Detection
• Pseudo-label Induced Subspace Representation Learning for Robust Out-of-Distribution Detection
• Diffusion-Scheduled Denoising Autoencoders for Anomaly Detection in Tabular Data

Related

• How Does Unlabeled Data Provably Help Out-of-Distribution Detection?
• Understanding the Role of Self-Supervised Learning in Out-of-Distribution Detection Task
• Do We Really Need to Learn Representations from In-domain Data for Outlier Detection?
• Three Factors to Improve Out-of-Distribution Detection
• Learning Representations for Outlier Detection on a Budget