Concurrent Misclassification and Out-of-Distribution Detection for Semantic Segmentation via Energy-Based Normalizing Flow
This addresses robustness issues in semantic segmentation for practical applications like autonomous driving, though it is incremental as it builds on existing normalizing flow and energy-based methods.
The paper tackles the problem of unreliable confidence scores in semantic segmentation under distribution shifts and out-of-distribution classes by proposing FlowEneDet, a generative model based on normalizing flow and energy-based inputs, which achieves promising results on benchmarks like Cityscapes and FishyScapes without requiring retraining of existing models.
Recent semantic segmentation models accurately classify test-time examples that are similar to a training dataset distribution. However, their discriminative closed-set approach is not robust in practical data setups with distributional shifts and out-of-distribution (OOD) classes. As a result, the predicted probabilities can be very imprecise when used as confidence scores at test time. To address this, we propose a generative model for concurrent in-distribution misclassification (IDM) and OOD detection that relies on a normalizing flow framework. The proposed flow-based detector with an energy-based inputs (FlowEneDet) can extend previously deployed segmentation models without their time-consuming retraining. Our FlowEneDet results in a low-complexity architecture with marginal increase in the memory footprint. FlowEneDet achieves promising results on Cityscapes, Cityscapes-C, FishyScapes and SegmentMeIfYouCan benchmarks in IDM/OOD detection when applied to pretrained DeepLabV3+ and SegFormer semantic segmentation models.