Class Relevance Learning For Out-of-distribution Detection
This addresses the challenge of detecting unfamiliar classes in real-world model deployment, which is crucial for safety but incremental in approach.
The paper tackles the problem of out-of-distribution (OOD) detection in image classification by introducing a class relevance learning method that leverages interclass relationships, demonstrating superior performance over state-of-the-art alternatives on diverse datasets.
Image classification plays a pivotal role across diverse applications, yet challenges persist when models are deployed in real-world scenarios. Notably, these models falter in detecting unfamiliar classes that were not incorporated during classifier training, a formidable hurdle for safe and effective real-world model deployment, commonly known as out-of-distribution (OOD) detection. While existing techniques, like max logits, aim to leverage logits for OOD identification, they often disregard the intricate interclass relationships that underlie effective detection. This paper presents an innovative class relevance learning method tailored for OOD detection. Our method establishes a comprehensive class relevance learning framework, strategically harnessing interclass relationships within the OOD pipeline. This framework significantly augments OOD detection capabilities. Extensive experimentation on diverse datasets, encompassing generic image classification datasets (Near OOD and Far OOD datasets), demonstrates the superiority of our method over state-of-the-art alternatives for OOD detection.