Timezone: »

OOD Detection with Class Ratio Estimation
Mingtian Zhang · Andi Zhang · Tim Xiao · Yitong Sun · Steven McDonagh

Density-based Out-of-distribution (OOD) detection has recently been shown unreliable for the task of detecting OOD images. Various density ratio based approaches have achieved good empirical performance. However, these methods typically lack a principled probabilistic modelling explanation. In this work, we propose to unify density ratio based methods under a novel energy-based model framework that allows us to view the density ratio as the unnormalized density of an implicit semantic distribution. Further, we propose to directly estimate the density ratio of a data sample through class ratio estimation, which can achieve competitive OOD detection results without training any deep generative models. Our approach enables a simple yet effective path towards solving OOD detection problems in the image domain.

Author Information

Mingtian Zhang (UCL)
Andi Zhang (University of Cambridge)
Tim Xiao (University of Tuebingen)
Yitong Sun (Huawei Noah's Ark Lab)
Steven McDonagh (Huawei Noah's Ark Lab)

More from the Same Authors