Label Smoothed Embedding Hypothesis for Out-of-Distribution Detection
Lists
Only you see this
Abstract
Detecting out-of-distribution (OOD) examples is critical in many applications. We propose an unsupervised method to detect OOD samples using a $k$-NN density estimate with respect to a classification model's intermediate activations on in-distribution samples. We leverage a recent insight about label smoothing, which we call the {\it Label Smoothed Embedding Hypothesis}, and show that one of the implications is that the $k$-NN density estimator performs better as an OOD detection method both theoretically and empirically when the model is trained with label smoothing. Finally, we show that our proposal outperforms many OOD baselines and we also provide new finite-sample high-probability statistical results for $k$-NN density estimation's ability to detect OOD examples.
BibTeX
@inproceedings{bahri2022label,
author = {Dara Bahri and Heinrich Jiang and Yi Tay and Donald Metzler},
title = {{Label Smoothed Embedding Hypothesis for Out-of-Distribution Detection}},
booktitle = {International Conference on Learning Representations},
year = {2022},
url = {https://openreview.net/forum?id=YtdASzotUEW},
}