Timezone: »

 
Deep Classifiers with Label Noise Modeling and Distance Awareness
Vincent Fortuin · Mark Collier · Florian Wenzel · James Allingham · Jeremiah Liu · Dustin Tran · Balaji Lakshminarayanan · Jesse Berent · Rodolphe Jenatton · Effrosyni Kokiopoulou

@
in Bayesian Deep Learning »
Event URL: https://openreview.net/forum?id=mypagftXxGQ »

Uncertainty estimation in deep learning has recently emerged as a crucial area of interest to advance reliability and robustness in safety-critical applications. While there have been many proposed methods that either focus on distance-aware model uncertainties for out-of-distribution detection or on input-dependent label uncertainties for in-distribution calibration, both of these types of uncertainty are often necessary. In this work, we propose the HetSNGP method for jointly modeling the model and data uncertainty. We show that our proposed model affords a favorable combination between these two complementary types of uncertainty and thus outperforms the baseline methods on some challenging out-of-distribution datasets, including CIFAR-100C, Imagenet-C, and Imagenet-A. Moreover, we propose HetSNGP Ensemble, an ensembled version of our method which adds an additional type of uncertainty and also outperforms other ensemble baselines.

Author Information

Vincent Fortuin (ETH Zürich)
Mark Collier (Google)
Florian Wenzel (---)
James Allingham (University of Cambridge)
Jeremiah Liu (Google Research / Harvard)
Dustin Tran (Google Brain)
Balaji Lakshminarayanan (Google Brain)
Jesse Berent
Rodolphe Jenatton (Amazon)
Effrosyni Kokiopoulou (Google AI)

More from the Same Authors