Why Should we Combine Training and Post-Training Methods for Out-of-Distribution Detection?

Aristotelis-Angelos Papadopoulos; Nazim Shaikh; Mohammad Reza Rajati

Why Should we Combine Training and Post-Training Methods for Out-of-Distribution Detection?

Machine Learning 2019-12-09 v1 Computer Vision and Pattern Recognition Machine Learning

Authors: Aristotelis-Angelos Papadopoulos , Nazim Shaikh , Mohammad Reza Rajati

Abstract

Deep neural networks are known to achieve superior results in classification tasks. However, it has been recently shown that they are incapable to detect examples that are generated by a distribution which is different than the one they have been trained on since they are making overconfident prediction for Out-Of-Distribution (OOD) examples. OOD detection has attracted a lot of attention recently. In this paper, we review some of the most seminal recent algorithms in the OOD detection field, we divide those methods into training and post-training and we experimentally show how the combination of the former with the latter can achieve state-of-the-art results in the OOD detection task.

Keywords

out-of-distribution detection deep neural networks training methods post-training methods state-of-the-art

Cite

@article{arxiv.1912.03133,
  title  = {Why Should we Combine Training and Post-Training Methods for Out-of-Distribution Detection?},
  author = {Aristotelis-Angelos Papadopoulos and Nazim Shaikh and Mohammad Reza Rajati},
  journal= {arXiv preprint arXiv:1912.03133},
  year   = {2019}
}

Comments

Preprint, 9 pages. arXiv admin note: text overlap with arXiv:1906.03509

Why Should we Combine Training and Post-Training Methods for Out-of-Distribution Detection?

Abstract

Keywords

Cite

Comments

Related papers