Supervision Adaptation Balancing In-Distribution Generalization and Out-of-Distribution Detection.

Zhao, Z; Cao, L; Lin, K-Y

Supervision Adaptation Balancing In-Distribution Generalization and Out-of-Distribution Detection.

Zhao, Z Cao, L

Lin, K-Y

Permalink

Publisher:: IEEE COMPUTER SOC
Publication Type:: Journal Article
Citation:: IEEE Trans Pattern Anal Mach Intell, 2023, 45, (12), pp. 15743-15758
Issue Date:: 2023-12

Closed Access

	Filename	Description	Size
	Supervision_Adaptation_Balancing_In-Distribution_Generalization_and_Out-of-Distribution_Detection.pdf	Published version	2.61 MB	Adobe PDF	View/Open

Copyright Clearance Process

Recently Added
In Progress
Closed Access

This item is closed access and not available.

Full metadata record

Field	Value	Language
dc.contributor.author	Zhao, Z
dc.contributor.author	Cao, L https://orcid.org/0000-0003-1562-9429
dc.contributor.author	Lin, K-Y
dc.date.accessioned	2024-04-12T04:02:10Z
dc.date.available	2024-04-12T04:02:10Z
dc.date.issued	2023-12
dc.identifier.citation	IEEE Trans Pattern Anal Mach Intell, 2023, 45, (12), pp. 15743-15758
dc.identifier.issn	0162-8828
dc.identifier.issn	1939-3539
dc.identifier.uri	http://hdl.handle.net/10453/177819
dc.description.abstract	The discrepancy between in-distribution (ID) and out-of-distribution (OOD) samples can lead to distributional vulnerability in deep neural networks, which can subsequently lead to high-confidence predictions for OOD samples. This is mainly due to the absence of OOD samples during training, which fails to constrain the network properly. To tackle this issue, several state-of-the-art methods include adding extra OOD samples to training and assign them with manually-defined labels. However, this practice can introduce unreliable labeling, negatively affecting ID classification. The distributional vulnerability presents a critical challenge for non-IID deep learning, which aims for OOD-tolerant ID classification by balancing ID generalization and OOD detection. In this paper, we introduce a novel supervision adaptation approach to generate adaptive supervision information for OOD samples, making them more compatible with ID samples. First, we measure the dependency between ID samples and their labels using mutual information, revealing that the supervision information can be represented in terms of negative probabilities across all classes. Second, we investigate data correlations between ID and OOD samples by solving a series of binary regression problems, with the goal of refining the supervision information for more distinctly separable ID classes. Our extensive experiments on four advanced network architectures, two ID datasets, and eleven diversified OOD datasets demonstrate the efficacy of our supervision adaptation approach in improving both ID classification and OOD detection capabilities.
dc.format	Print-Electronic
dc.language	eng
dc.publisher	IEEE COMPUTER SOC
dc.relation	http://purl.org/au-research/grants/arc/DP190101079
dc.relation	http://purl.org/au-research/grants/arc/FT190100734
dc.relation.ispartof	IEEE Trans Pattern Anal Mach Intell
dc.relation.isbasedon	10.1109/TPAMI.2023.3321869
dc.rights	info:eu-repo/semantics/closedAccess
dc.subject	0801 Artificial Intelligence and Image Processing, 0806 Information Systems, 0906 Electrical and Electronic Engineering
dc.subject.classification	Artificial Intelligence & Image Processing
dc.subject.classification	4603 Computer vision and multimedia computation
dc.subject.classification	4611 Machine learning
dc.title	Supervision Adaptation Balancing In-Distribution Generalization and Out-of-Distribution Detection.
dc.type	Journal Article
utslib.citation.volume	45
utslib.location.activity	United States
utslib.for	0801 Artificial Intelligence and Image Processing
utslib.for	0806 Information Systems
utslib.for	0906 Electrical and Electronic Engineering
pubs.organisational-group	University of Technology Sydney
pubs.organisational-group	University of Technology Sydney/Faculty of Engineering and Information Technology
pubs.organisational-group	University of Technology Sydney/Strength - AAI - Advanced Analytics Institute Research Centre
utslib.copyright.status	closed_access	*
dc.date.updated	2024-04-12T04:02:08Z
pubs.issue	12
pubs.publication-status	Published
pubs.volume	45
utslib.citation.issue	12

Abstract:

The discrepancy between in-distribution (ID) and out-of-distribution (OOD) samples can lead to distributional vulnerability in deep neural networks, which can subsequently lead to high-confidence predictions for OOD samples. This is mainly due to the absence of OOD samples during training, which fails to constrain the network properly. To tackle this issue, several state-of-the-art methods include adding extra OOD samples to training and assign them with manually-defined labels. However, this practice can introduce unreliable labeling, negatively affecting ID classification. The distributional vulnerability presents a critical challenge for non-IID deep learning, which aims for OOD-tolerant ID classification by balancing ID generalization and OOD detection. In this paper, we introduce a novel supervision adaptation approach to generate adaptive supervision information for OOD samples, making them more compatible with ID samples. First, we measure the dependency between ID samples and their labels using mutual information, revealing that the supervision information can be represented in terms of negative probabilities across all classes. Second, we investigate data correlations between ID and OOD samples by solving a series of binary regression problems, with the goal of refining the supervision information for more distinctly separable ID classes. Our extensive experiments on four advanced network architectures, two ID datasets, and eleven diversified OOD datasets demonstrate the efficacy of our supervision adaptation approach in improving both ID classification and OOD detection capabilities.

Please use this identifier to cite or link to this item:

http://hdl.handle.net/10453/177819