Hybrid Models for Open Set Recognition

Abstract

Open set recognition requires a classifier to detect samples not belonging to any of the classes in its training set. Existing methods fit a probability distribution to the training samples on their embedding space and detect outliers according to this distribution. The embedding space is often obtained from a discriminative classifier. However, such discriminative representation focuses only on known classes, which may not be critical for distinguishing the unknown classes. We argue that the representation space should be jointly learned from the inlier classifier and the density estimator (served as an outlier detector). We propose the OpenHybrid framework, which is composed of an encoder to encode the input data into a joint embedding space, a classifier to classify samples to inlier classes, and a flow-based density estimator to detect whether a sample belongs to the unknown category. A typical problem of existing flow-based models is that they may assign a higher likelihood to outliers. However, we empirically observe that such an issue does not occur in our experiments when learning a joint representation for discriminative and generative components. Experiments on standard open set benchmarks also reveal that an end-to-end trained OpenHybrid model significantly outperforms state-of-the-art methods and flow-based baselines.

arXiv comment16 pages, 4 figures

References in corpus (14)

• NICE: Non-linear Independent Components Estimation
• Recent Advances in Open Set Recognition: A Survey
• Density estimation using Real NVP
• Deep Anomaly Detection with Outlier Exposure
• Residual Flows for Invertible Generative Modeling
• Invertible Residual Networks
• Input complexity and out-of-distribution detection with likelihood-based generative models
• Training Confidence-calibrated Classifiers for Detecting Out-of-Distribution Samples
• Generative OpenMax for Multi-Class Open Set Classification
• Do Deep Generative Models Know What They Don't Know?
• Detecting Out-of-Distribution Inputs to Deep Generative Models Using Typicality
• Out-of-distribution Detection in Classifiers via Generation
• DOC: Deep Open Classification of Text Documents
• Likelihood Assignment for Out-of-Distribution Inputs in Deep Generative Models is Sensitive to Prior Distribution Choice

Cited by in corpus (5)

• Contrastive Training for Improved Out-of-Distribution Detection
• Perfect density models cannot guarantee anomaly detection
• Task-agnostic Continual Learning with Hybrid Probabilistic Models
• Unsupervised Anomaly Detection From Semantic Similarity Scores
• Revealing the Distributional Vulnerability of Discriminators by Implicit Generators