Multi-Label Out-of-Distribution Detection with Spectral Normalized Joint Energy
This work addresses the challenge of reliable OOD detection for multi-label machine learning models, which is incremental as it builds on existing OOD detection methods but extends them to the under-explored multi-label setting.
The paper tackles the problem of multi-label out-of-distribution detection, which has been limited in prior research, by introducing Spectral Normalized Joint Energy (SNoJoE), achieving 11% and 54% relative reductions in FPR95 on OOD datasets compared to previous top performances.
In today's interconnected world, achieving reliable out-of-distribution (OOD) detection poses a significant challenge for machine learning models. While numerous studies have introduced improved approaches for multi-class OOD detection tasks, the investigation into multi-label OOD detection tasks has been notably limited. We introduce Spectral Normalized Joint Energy (SNoJoE), a method that consolidates label-specific information across multiple labels through the theoretically justified concept of an energy-based function. Throughout the training process, we employ spectral normalization to manage the model's feature space, thereby enhancing model efficacy and generalization, in addition to bolstering robustness. Our findings indicate that the application of spectral normalization to joint energy scores notably amplifies the model's capability for OOD detection. We perform OOD detection experiments utilizing PASCAL-VOC as the in-distribution dataset and ImageNet-22K or Texture as the out-of-distribution datasets. Our experimental results reveal that, in comparison to prior top performances, SNoJoE achieves 11% and 54% relative reductions in FPR95 on the respective OOD datasets, thereby defining the new state of the art in this field of study.