CCAC

2000년 도입 · 논문 2편에서 사용

Confidence Calibration with an Auxiliary Class, or CCAC, is a post-hoc confidence calibration method for DNN classifiers on OOD datasets. The key feature of CCAC is an auxiliary class in the calibration model which separates mis-classified samples from correctly classified ones, thus effectively mitigating the target DNN’s being confidently wrong. It also reduces the number of free parameters in CCAC to reduce free parameters and facilitate transfer to a new unseen dataset.

출처: Calibrating Deep Neural Network Classifiers on Out-of-Distribution Datasets

소개 논문: Calibrating Deep Neural Network Classifiers on Out-of-Distribution Datasets