<ul><li>ProHOC is a framework for detecting and classifying out-of-distribution (OOD) samples in a class hierarchy.</li><li>The approach leverages a probabilistic model that uses networks trained for in-distribution (ID) classification at multiple hierarchy depths.</li><li>Experiments conducted on three datasets with predefined class hierarchies demonstrate the effectiveness of the method.</li><li>The code for ProHOC is available at https://github.com/walline/prohoc.</li></ul>

ProHOC: Probabilistic Hierarchical Out-of-Distribution Classification via Multi-Depth Networks

Discover more