Resultant: Incremental Effectiveness on Likelihood for Unsupervised Out-of-Distribution Detection

TL;DR

Resultant is a new method that enhances likelihood-based unsupervised out-of-distribution detection by combining techniques to consistently outperform or match likelihood performance across diverse benchmarks.

Contribution

The paper introduces Resultant, a novel approach that improves likelihood-based U-OOD detection by integrating two techniques, ensuring incremental effectiveness across various datasets.

Findings

Achieves state-of-the-art U-OOD detection performance

Maintains incremental effectiveness on likelihood in diverse benchmarks

Combines post-hoc prior and dataset entropy-mutual calibration techniques

Abstract

Unsupervised out-of-distribution (U-OOD) detection is to identify OOD data samples with a detector trained solely on unlabeled in-distribution (ID) data. The likelihood function estimated by a deep generative model (DGM) could be a natural detector, but its performance is limited in some popular "hard" benchmarks, such as FashionMNIST (ID) vs. MNIST (OOD). Recent studies have developed various detectors based on DGMs to move beyond likelihood. However, despite their success on "hard" benchmarks, most of them struggle to consistently surpass or match the performance of likelihood on some "non-hard" cases, such as SVHN (ID) vs. CIFAR10 (OOD) where likelihood could be a nearly perfect detector. Therefore, we appeal for more attention to incremental effectiveness on likelihood, i.e., whether a method could always surpass or at least match the performance of likelihood in U-OOD detection. We first investigate the likelihood of variational DGMs and find its detection performance could be improved in two directions: i) alleviating latent distribution mismatch, and ii) calibrating the dataset entropy-mutual integration. Then, we apply two techniques for each direction, specifically post-hoc prior and dataset entropy-mutual calibration. The final method, named Resultant, combines these two directions for better incremental effectiveness compared to either technique alone. Experimental results demonstrate that the Resultant could be a new state-of-the-art U-OOD detector while maintaining incremental effectiveness on likelihood in a wide range of tasks.