Detecting semantic anomalies
This work addresses the need for more realistic OOD detection benchmarks in computer vision, though it is incremental in refining existing approaches.
The paper tackles the problem of out-of-distribution (OOD) detection by arguing that current benchmarks lack practical relevance and proposing semantic anomaly detection for object recognition, resulting in improved performance with multi-task learning and generalization benefits.
We critically appraise the recent interest in out-of-distribution (OOD) detection and question the practical relevance of existing benchmarks. While the currently prevalent trend is to consider different datasets as OOD, we argue that out-distributions of practical interest are ones where the distinction is semantic in nature for a specified context, and that evaluative tasks should reflect this more closely. Assuming a context of object recognition, we recommend a set of benchmarks, motivated by practical applications. We make progress on these benchmarks by exploring a multi-task learning based approach, showing that auxiliary objectives for improved semantic awareness result in improved semantic anomaly detection, with accompanying generalization benefits.