English

Revisiting Confidence Estimation: Towards Reliable Failure Prediction

Computer Vision and Pattern Recognition 2024-03-06 v1 Machine Learning

Abstract

Reliable confidence estimation is a challenging yet fundamental requirement in many risk-sensitive applications. However, modern deep neural networks are often overconfident for their incorrect predictions, i.e., misclassified samples from known classes, and out-of-distribution (OOD) samples from unknown classes. In recent years, many confidence calibration and OOD detection methods have been developed. In this paper, we find a general, widely existing but actually-neglected phenomenon that most confidence estimation methods are harmful for detecting misclassification errors. We investigate this problem and reveal that popular calibration and OOD detection methods often lead to worse confidence separation between correctly classified and misclassified examples, making it difficult to decide whether to trust a prediction or not. Finally, we propose to enlarge the confidence gap by finding flat minima, which yields state-of-the-art failure prediction performance under various settings including balanced, long-tailed, and covariate-shift classification scenarios. Our study not only provides a strong baseline for reliable confidence estimation but also acts as a bridge between understanding calibration, OOD detection, and failure prediction. The code is available at \url{https://github.com/Impression2805/FMFP}.

Keywords

Cite

@article{arxiv.2403.02886,
  title  = {Revisiting Confidence Estimation: Towards Reliable Failure Prediction},
  author = {Fei Zhu and Xu-Yao Zhang and Zhen Cheng and Cheng-Lin Liu},
  journal= {arXiv preprint arXiv:2403.02886},
  year   = {2024}
}

Comments

Accepted by IEEE TPAMI. arXiv admin note: text overlap with arXiv:2303.02970; text overlap with arXiv:2007.01458 by other authors

R2 v1 2026-06-28T15:09:40.337Z