FROB: Few-Shot ROBust Model for Joint Classification and Out-of-Distribution Detection

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract / Description of output

Classification and Out-of-Distribution (OoD) detection in the few-shot setting remain challenging aims, but are important for devising critical systems in security where samples are limited. OoD detection requires that classifiers are aware of when they do not know and avoid setting high confidence to OoD samples away from the training data distribution. To address such limitations, we propose the Few-shot ROBust (FROB) model with its key contributions being (a) the joint classification and few-shot OoD detection, (b) the sample generation on the boundary of the support of the normal class distribution, and (c) the incorporation of the learned distribution boundary as OoD data for contrastive negative training. FROB finds the boundary of the support of the normal class distribution, and uses it to improve the few-shot OoD detection performance. We propose a self-supervised learning methodology for sample generation on the normal class distribution confidence boundary based on generative and discriminative models, including classification. FROB implicitly generates adversarial samples, and forces samples from OoD, including our boundary, to be less confident by the classifier. By including the learned boundary, FROB reduces the threshold linked to the model’s few-shot robustness in the number of few-shots, and maintains the OoD performance approximately constant, independent of the number of few-shots. The low- and few-shot robustness evaluation of FROB on different image datasets and on One-Class Classification (OCC) data shows that FROB achieves competitive performance and outperforms baselines in terms of robustness to the OoD few-shot population and variability.
Original languageEnglish
Title of host publicationMachine Learning and Knowledge Discovery in Databases
Subtitle of host publicationEuropean Conference, ECML PKDD 2022, Grenoble, France, September 19–23, 2022, Proceedings, Part III
PublisherSpringer
Pages137-153
Volume13715
ISBN (Electronic)978-3-031-26409-2
ISBN (Print)978-3-031-26408-5
DOIs
Publication statusE-pub ahead of print - 17 Mar 2023
EventEuropean Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases 2022 - Grenoble, France
Duration: 19 Sept 202223 Sept 2022
https://2022.ecmlpkdd.org/

Publication series

NameLecture Notes in Computer Science
PublisherSpringer
Volume13715
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

ConferenceEuropean Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases 2022
Abbreviated titleECML PKDD 2022
Country/TerritoryFrance
CityGrenoble
Period19/09/2223/09/22
Internet address

Keywords / Materials (for Non-textual outputs)

  • Few-shot anomaly detection
  • Out-of-Distribution detection

Fingerprint

Dive into the research topics of 'FROB: Few-Shot ROBust Model for Joint Classification and Out-of-Distribution Detection'. Together they form a unique fingerprint.

Cite this