Post-Hoc Uncertainty Calibration for Domain Drift Scenarios

MCML Authors

Christian Tomani

* Former Member

→ Group Daniel Cremers
 Computer Vision & Artificial Intelligence

Daniel Cremers

Prof. Dr.

Director

Computer Vision & Artificial Intelligence

Abstract

We address the problem of uncertainty calibration. While standard deep neural networks typically yield uncalibrated predictions, calibrated confidence scores that are representative of the true likelihood of a prediction can be achieved using post-hoc calibration methods. However, to date, the focus of these approaches has been on in-domain calibration. Our contribution is two-fold. First, we show that existing post-hoc calibration methods yield highly over-confident predictions under domain shift. Second, we introduce a simple strategy where perturbations are applied to samples in the validation set before performing the post-hoc calibration step. In extensive experiments, we demonstrate that this perturbation step results in substantially better calibration under domain shift on a wide range of architectures and modelling tasks.

inproceedings TGE+21

---

CVPR 2021

IEEE/CVF Conference on Computer Vision and Pattern Recognition. Virtual, Jun 19-25, 2021.

---

Authors

C. Tomani • S. Gruber • M. E. Erdem • D. Cremers • F. Buettner

---

Links

DOI

---

In Collaboration

 Siemens AG

---

Research Area

 B1 | Computer Vision