A Simple and Effective Baseline for Out-of-Distribution Detection using AbstentionDownload PDF

28 Sept 2020, 15:52 (edited 05 Mar 2021)ICLR 2021 Conference Blind SubmissionReaders: Everyone
  • Reviewed Version (pdf): https://openreview.net/references/pdf?id=6VlGUgDcN
  • Keywords: deep learning, out-of-distribution detection
  • Abstract: Refraining from confidently predicting when faced with categories of inputs different from those seen during training is an important requirement for the safe deployment of deep learning systems. While simple to state, this has been a particularly challenging problem in deep learning, where models often end up making overconfident predictions in such situations. In this work we present a simple, but highly effective approach to deal with out-of-distribution detection that uses the principle of abstention: when encountering a sample from an unseen class, the desired behavior is to abstain from predicting. Our approach uses a network with an extra abstention class and is trained on a dataset that is augmented with an uncurated set that consists of a large number of out-of-distribution (OoD) samples that are assigned the label of the abstention class; the model is then trained to learn an effective discriminator between in and out-of-distribution samples. We compare this relatively simple approach against a wide variety of more complex methods that have been proposed both for out-of-distribution detection as well as uncertainty modeling in deep learning, and empirically demonstrate its effectiveness on a wide variety of of benchmarks and deep architectures for image recognition and text classification, often outperforming existing approaches by significant margins. Given the simplicity and effectiveness of this method, we propose that this approach be used as a new additional baseline for future work in this domain.
  • One-sentence Summary: Deep neural networks augmented with an extra abstention class and trained on in and out-of-distribution data show strong out-of-detection performance, often exceeding existing state-of-the-art.
  • Supplementary Material: zip
  • Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics
15 Replies