SimLabel: Consistency-Guided OOD Detection with Pretrained Vision-Language Models Article Swipe

PDF

Shu Zou , Xinyu Tian , Qinyu Zhao , Zhaoyuan Yang , Jing Zhang ·

YOU? · · 2025 · Open Access · · DOI: https://doi.org/10.48550/arxiv.2501.11485

Detecting out-of-distribution (OOD) data is crucial in real-world machine learning applications, particularly in safety-critical domains. Existing methods often leverage language information from vision-language models (VLMs) to enhance OOD detection by improving confidence estimation through rich class-wise text information. However, when building OOD detection score upon on in-distribution (ID) text-image affinity, existing works either focus on each ID class or whole ID label sets, overlooking inherent ID classes' connection. We find that the semantic information across different ID classes is beneficial for effective OOD detection. We thus investigate the ability of image-text comprehension among different semantic-related ID labels in VLMs and propose a novel post-hoc strategy called SimLabel. SimLabel enhances the separability between ID and OOD samples by establishing a more robust image-class similarity metric that considers consistency over a set of similar class labels. Extensive experiments demonstrate the superior performance of SimLabel on various zero-shot OOD detection benchmarks. The proposed model is also extended to various VLM-backbones, demonstrating its good generalization ability. Our demonstration and implementation codes are available at: https://github.com/ShuZou-1/SimLabel.

Related Topics

Computer Science

Artificial Intelligence

Concepts

Consistency (knowledge bases) Computer science Artificial intelligence Natural language processing Language model

Metadata

Type: preprint
Language: en
Landing Page: http://arxiv.org/abs/2501.11485
PDF: https://arxiv.org/pdf/2501.11485
OA Status: green
Cited By: 1
Related Works: 10
OpenAlex ID: https://openalex.org/W4406744325

All OpenAlex metadata

Raw OpenAlex JSON

OpenAlex ID: https://openalex.org/W4406744325

Canonical identifier for this work in OpenAlex
DOI: https://doi.org/10.48550/arxiv.2501.11485

Digital Object Identifier
Title: SimLabel: Consistency-Guided OOD Detection with Pretrained Vision-Language Models

Work title
Type: preprint

OpenAlex work type
Language: en

Primary language
Publication year: 2025

Year of publication
Publication date: 2025-01-20

Full publication date if available
Authors: Shu Zou, Xinyu Tian, Qinyu Zhao, Zhaoyuan Yang, Jing Zhang

List of authors in order
Landing page: https://arxiv.org/abs/2501.11485

Publisher landing page
PDF URL: https://arxiv.org/pdf/2501.11485

Direct link to full text PDF
Open access: Yes

Whether a free full text is available
OA status: green

Open access status per OpenAlex
OA URL: https://arxiv.org/pdf/2501.11485

Direct OA link when available
Concepts: Consistency (knowledge bases), Computer science, Artificial intelligence, Natural language processing, Language model

Top concepts (fields/topics) attached by OpenAlex
Cited by: 1

Total citation count in OpenAlex
Citations by year (recent): 2025: 1

Per-year citation counts (last 5 years)
Related works (count): 10

Other works algorithmically related by OpenAlex

Full payload

id	https://openalex.org/W4406744325
doi	https://doi.org/10.48550/arxiv.2501.11485
ids.doi	https://doi.org/10.48550/arxiv.2501.11485
ids.openalex	https://openalex.org/W4406744325
fwci
type	preprint
title	SimLabel: Consistency-Guided OOD Detection with Pretrained Vision-Language Models
biblio.issue
biblio.volume
biblio.last_page
biblio.first_page
topics[0].id	https://openalex.org/T12016
topics[0].field.id	https://openalex.org/fields/17
topics[0].field.display_name	Computer Science
topics[0].score	0.9807999730110168
topics[0].domain.id	https://openalex.org/domains/3
topics[0].domain.display_name	Physical Sciences
topics[0].subfield.id	https://openalex.org/subfields/1710
topics[0].subfield.display_name	Information Systems
topics[0].display_name	Web Data Mining and Analysis
is_xpac	False
apc_list
apc_paid
concepts[0].id	https://openalex.org/C2776436953
concepts[0].level	2
concepts[0].score	0.7179067730903625
concepts[0].wikidata	https://www.wikidata.org/wiki/Q5163215
concepts[0].display_name	Consistency (knowledge bases)
concepts[1].id	https://openalex.org/C41008148
concepts[1].level	0
concepts[1].score	0.555220901966095
concepts[1].wikidata	https://www.wikidata.org/wiki/Q21198
concepts[1].display_name	Computer science
concepts[2].id	https://openalex.org/C154945302
concepts[2].level	1
concepts[2].score	0.49289751052856445
concepts[2].wikidata	https://www.wikidata.org/wiki/Q11660
concepts[2].display_name	Artificial intelligence
concepts[3].id	https://openalex.org/C204321447
concepts[3].level	1
concepts[3].score	0.43384116888046265
concepts[3].wikidata	https://www.wikidata.org/wiki/Q30642
concepts[3].display_name	Natural language processing
concepts[4].id	https://openalex.org/C137293760
concepts[4].level	2
concepts[4].score	0.41355347633361816
concepts[4].wikidata	https://www.wikidata.org/wiki/Q3621696
concepts[4].display_name	Language model
keywords[0].id	https://openalex.org/keywords/consistency
keywords[0].score	0.7179067730903625
keywords[0].display_name	Consistency (knowledge bases)
keywords[1].id	https://openalex.org/keywords/computer-science
keywords[1].score	0.555220901966095
keywords[1].display_name	Computer science
keywords[2].id	https://openalex.org/keywords/artificial-intelligence
keywords[2].score	0.49289751052856445
keywords[2].display_name	Artificial intelligence
keywords[3].id	https://openalex.org/keywords/natural-language-processing
keywords[3].score	0.43384116888046265
keywords[3].display_name	Natural language processing
keywords[4].id	https://openalex.org/keywords/language-model
keywords[4].score	0.41355347633361816
keywords[4].display_name	Language model
language	en
locations[0].id	pmh:oai:arXiv.org:2501.11485
locations[0].is_oa	True
locations[0].source.id	https://openalex.org/S4306400194
locations[0].source.issn
locations[0].source.type	repository
locations[0].source.is_oa	True
locations[0].source.issn_l
locations[0].source.is_core	False
locations[0].source.is_in_doaj	False
locations[0].source.display_name	arXiv (Cornell University)
locations[0].source.host_organization	https://openalex.org/I205783295
locations[0].source.host_organization_name	Cornell University
locations[0].source.host_organization_lineage	https://openalex.org/I205783295
locations[0].license
locations[0].pdf_url	https://arxiv.org/pdf/2501.11485
locations[0].version	submittedVersion
locations[0].raw_type	text
locations[0].license_id
locations[0].is_accepted	False
locations[0].is_published	False
locations[0].raw_source_name
locations[0].landing_page_url	http://arxiv.org/abs/2501.11485
locations[1].id	doi:10.48550/arxiv.2501.11485
locations[1].is_oa	True
locations[1].source.id	https://openalex.org/S4306400194
locations[1].source.issn
locations[1].source.type	repository
locations[1].source.is_oa	True
locations[1].source.issn_l
locations[1].source.is_core	False
locations[1].source.is_in_doaj	False
locations[1].source.display_name	arXiv (Cornell University)
locations[1].source.host_organization	https://openalex.org/I205783295
locations[1].source.host_organization_name	Cornell University
locations[1].source.host_organization_lineage	https://openalex.org/I205783295
locations[1].license
locations[1].pdf_url
locations[1].version
locations[1].raw_type	article
locations[1].license_id
locations[1].is_accepted	False
locations[1].is_published
locations[1].raw_source_name
locations[1].landing_page_url	https://doi.org/10.48550/arxiv.2501.11485
indexed_in	arxiv, datacite
authorships[0].author.id	https://openalex.org/A5101292475
authorships[0].author.orcid
authorships[0].author.display_name	Shu Zou
authorships[0].author_position	first
authorships[0].raw_author_name	Zou, Shu
authorships[0].is_corresponding	False
authorships[1].author.id	https://openalex.org/A5029944382
authorships[1].author.orcid	https://orcid.org/0000-0003-1247-6076
authorships[1].author.display_name	Xinyu Tian
authorships[1].author_position	middle
authorships[1].raw_author_name	Tian, Xinyu
authorships[1].is_corresponding	False
authorships[2].author.id	https://openalex.org/A5101153090
authorships[2].author.orcid	https://orcid.org/0000-0003-0245-1676
authorships[2].author.display_name	Qinyu Zhao
authorships[2].author_position	middle
authorships[2].raw_author_name	Zhao, Qinyu
authorships[2].is_corresponding	False
authorships[3].author.id	https://openalex.org/A5103063628
authorships[3].author.orcid	https://orcid.org/0009-0007-0294-4741
authorships[3].author.display_name	Zhaoyuan Yang
authorships[3].author_position	middle
authorships[3].raw_author_name	Yang, Zhaoyuan
authorships[3].is_corresponding	False
authorships[4].author.id	https://openalex.org/A5101681014
authorships[4].author.orcid	https://orcid.org/0000-0003-2019-225X
authorships[4].author.display_name	Jing Zhang
authorships[4].author_position	last
authorships[4].raw_author_name	Zhang, Jing
authorships[4].is_corresponding	False
has_content.pdf	False
has_content.grobid_xml	False
is_paratext	False
open_access.is_oa	True
open_access.oa_url	https://arxiv.org/pdf/2501.11485
open_access.oa_status	green
open_access.any_repository_has_fulltext	False
created_date	2025-10-10T00:00:00
display_name	SimLabel: Consistency-Guided OOD Detection with Pretrained Vision-Language Models
has_fulltext	False
is_retracted	False
updated_date	2025-11-06T06:51:31.235846
primary_topic.id	https://openalex.org/T12016
primary_topic.field.id	https://openalex.org/fields/17
primary_topic.field.display_name	Computer Science
primary_topic.score	0.9807999730110168
primary_topic.domain.id	https://openalex.org/domains/3
primary_topic.domain.display_name	Physical Sciences
primary_topic.subfield.id	https://openalex.org/subfields/1710
primary_topic.subfield.display_name	Information Systems
primary_topic.display_name	Web Data Mining and Analysis
related_works	https://openalex.org/W4391375266, https://openalex.org/W2899084033, https://openalex.org/W2748952813, https://openalex.org/W2390279801, https://openalex.org/W4391913857, https://openalex.org/W2358668433, https://openalex.org/W4396701345, https://openalex.org/W2376932109, https://openalex.org/W2001405890, https://openalex.org/W3204019825
cited_by_count	1
counts_by_year[0].year	2025
counts_by_year[0].cited_by_count	1
locations_count	2
best_oa_location.id	pmh:oai:arXiv.org:2501.11485
best_oa_location.is_oa	True
best_oa_location.source.id	https://openalex.org/S4306400194
best_oa_location.source.issn
best_oa_location.source.type	repository
best_oa_location.source.is_oa	True
best_oa_location.source.issn_l
best_oa_location.source.is_core	False
best_oa_location.source.is_in_doaj	False
best_oa_location.source.display_name	arXiv (Cornell University)
best_oa_location.source.host_organization	https://openalex.org/I205783295
best_oa_location.source.host_organization_name	Cornell University
best_oa_location.source.host_organization_lineage	https://openalex.org/I205783295
best_oa_location.license
best_oa_location.pdf_url	https://arxiv.org/pdf/2501.11485
best_oa_location.version	submittedVersion
best_oa_location.raw_type	text
best_oa_location.license_id
best_oa_location.is_accepted	False
best_oa_location.is_published	False
best_oa_location.raw_source_name
best_oa_location.landing_page_url	http://arxiv.org/abs/2501.11485
primary_location.id	pmh:oai:arXiv.org:2501.11485
primary_location.is_oa	True
primary_location.source.id	https://openalex.org/S4306400194
primary_location.source.issn
primary_location.source.type	repository
primary_location.source.is_oa	True
primary_location.source.issn_l
primary_location.source.is_core	False
primary_location.source.is_in_doaj	False
primary_location.source.display_name	arXiv (Cornell University)
primary_location.source.host_organization	https://openalex.org/I205783295
primary_location.source.host_organization_name	Cornell University
primary_location.source.host_organization_lineage	https://openalex.org/I205783295
primary_location.license
primary_location.pdf_url	https://arxiv.org/pdf/2501.11485
primary_location.version	submittedVersion
primary_location.raw_type	text
primary_location.license_id
primary_location.is_accepted	False
primary_location.is_published	False
primary_location.raw_source_name
primary_location.landing_page_url	http://arxiv.org/abs/2501.11485
publication_date	2025-01-20
publication_year	2025
referenced_works_count	0
abstract_inverted_index.a	101, 118, 128
abstract_inverted_index.ID	56, 60, 65, 76, 95, 112
abstract_inverted_index.We	68, 84
abstract_inverted_index.by	29, 116
abstract_inverted_index.in	6, 12, 97
abstract_inverted_index.is	4, 78, 151
abstract_inverted_index.of	89, 130, 140
abstract_inverted_index.on	45, 54, 142
abstract_inverted_index.or	58
abstract_inverted_index.to	25, 154
abstract_inverted_index.OOD	27, 41, 82, 114, 145
abstract_inverted_index.Our	162
abstract_inverted_index.The	148
abstract_inverted_index.and	99, 113, 164
abstract_inverted_index.are	167
abstract_inverted_index.at:	169
abstract_inverted_index.for	80
abstract_inverted_index.its	158
abstract_inverted_index.set	129
abstract_inverted_index.the	71, 87, 109, 137
abstract_inverted_index.(ID)	47
abstract_inverted_index.VLMs	98
abstract_inverted_index.also	152
abstract_inverted_index.data	3
abstract_inverted_index.each	55
abstract_inverted_index.find	69
abstract_inverted_index.from	21
abstract_inverted_index.good	159
abstract_inverted_index.more	119
abstract_inverted_index.over	127
abstract_inverted_index.rich	34
abstract_inverted_index.text	36
abstract_inverted_index.that	70, 124
abstract_inverted_index.thus	85
abstract_inverted_index.upon	44
abstract_inverted_index.when	39
abstract_inverted_index.(OOD)	2
abstract_inverted_index.among	92
abstract_inverted_index.class	57, 132
abstract_inverted_index.codes	166
abstract_inverted_index.focus	53
abstract_inverted_index.label	61
abstract_inverted_index.model	150
abstract_inverted_index.novel	102
abstract_inverted_index.often	17
abstract_inverted_index.score	43
abstract_inverted_index.sets,	62
abstract_inverted_index.whole	59
abstract_inverted_index.works	51
abstract_inverted_index.(VLMs)	24
abstract_inverted_index.across	74
abstract_inverted_index.called	105
abstract_inverted_index.either	52
abstract_inverted_index.labels	96
abstract_inverted_index.metric	123
abstract_inverted_index.models	23
abstract_inverted_index.robust	120
abstract_inverted_index.ability	88
abstract_inverted_index.between	111
abstract_inverted_index.classes	77
abstract_inverted_index.crucial	5
abstract_inverted_index.enhance	26
abstract_inverted_index.labels.	133
abstract_inverted_index.machine	8
abstract_inverted_index.methods	16
abstract_inverted_index.propose	100
abstract_inverted_index.samples	115
abstract_inverted_index.similar	131
abstract_inverted_index.through	33
abstract_inverted_index.various	143, 155
abstract_inverted_index.Existing	15
abstract_inverted_index.However,	38
abstract_inverted_index.SimLabel	107, 141
abstract_inverted_index.ability.	161
abstract_inverted_index.building	40
abstract_inverted_index.classes'	66
abstract_inverted_index.domains.	14
abstract_inverted_index.enhances	108
abstract_inverted_index.existing	50
abstract_inverted_index.extended	153
abstract_inverted_index.inherent	64
abstract_inverted_index.language	19
abstract_inverted_index.learning	9
abstract_inverted_index.leverage	18
abstract_inverted_index.post-hoc	103
abstract_inverted_index.proposed	149
abstract_inverted_index.semantic	72
abstract_inverted_index.strategy	104
abstract_inverted_index.superior	138
abstract_inverted_index.Detecting	0
abstract_inverted_index.Extensive	134
abstract_inverted_index.SimLabel.	106
abstract_inverted_index.affinity,	49
abstract_inverted_index.available	168
abstract_inverted_index.considers	125
abstract_inverted_index.detection	28, 42, 146
abstract_inverted_index.different	75, 93
abstract_inverted_index.effective	81
abstract_inverted_index.improving	30
abstract_inverted_index.zero-shot	144
abstract_inverted_index.beneficial	79
abstract_inverted_index.class-wise	35
abstract_inverted_index.confidence	31
abstract_inverted_index.detection.	83
abstract_inverted_index.estimation	32
abstract_inverted_index.image-text	90
abstract_inverted_index.real-world	7
abstract_inverted_index.similarity	122
abstract_inverted_index.text-image	48
abstract_inverted_index.benchmarks.	147
abstract_inverted_index.connection.	67
abstract_inverted_index.consistency	126
abstract_inverted_index.demonstrate	136
abstract_inverted_index.experiments	135
abstract_inverted_index.image-class	121
abstract_inverted_index.information	20, 73
abstract_inverted_index.investigate	86
abstract_inverted_index.overlooking	63
abstract_inverted_index.performance	139
abstract_inverted_index.establishing	117
abstract_inverted_index.information.	37
abstract_inverted_index.particularly	11
abstract_inverted_index.separability	110
abstract_inverted_index.applications,	10
abstract_inverted_index.comprehension	91
abstract_inverted_index.demonstrating	157
abstract_inverted_index.demonstration	163
abstract_inverted_index.VLM-backbones,	156
abstract_inverted_index.generalization	160
abstract_inverted_index.implementation	165
abstract_inverted_index.in-distribution	46
abstract_inverted_index.safety-critical	13
abstract_inverted_index.vision-language	22
abstract_inverted_index.semantic-related	94
abstract_inverted_index.out-of-distribution	1
abstract_inverted_index.https://github.com/ShuZou-1/SimLabel.	170
cited_by_percentile_year
countries_distinct_count	0
institutions_distinct_count	5
citation_normalized_percentile