MultiADS: Defect-aware Supervision for Multi-type Anomaly Detection and Segmentation in Zero-Shot Learning Article Swipe
Ylli Sadikaj
,
Hongkuan Zhou
,
Lavdim Halilaj
,
Stefan Schmid
,
Steffen Staab
,
Claudia Plant
·
YOU?
·
· 2025
· Open Access
·
· DOI: https://doi.org/10.48550/arxiv.2504.06740
YOU?
·
· 2025
· Open Access
·
· DOI: https://doi.org/10.48550/arxiv.2504.06740
Precise optical inspection in industrial applications is crucial for minimizing scrap rates and reducing the associated costs. Besides merely detecting if a product is anomalous or not, it is crucial to know the distinct type of defect, such as a bent, cut, or scratch. The ability to recognize the "exact" defect type enables automated treatments of the anomalies in modern production lines. Current methods are limited to solely detecting whether a product is defective or not without providing any insights on the defect type, nevertheless detecting and identifying multiple defects. We propose MultiADS, a zero-shot learning approach, able to perform Multi-type Anomaly Detection and Segmentation. The architecture of MultiADS comprises CLIP and extra linear layers to align the visual- and textual representation in a joint feature space. To the best of our knowledge, our proposal, is the first approach to perform a multi-type anomaly segmentation task in zero-shot learning. Contrary to the other baselines, our approach i) generates specific anomaly masks for each distinct defect type, ii) learns to distinguish defect types, and iii) simultaneously identifies multiple defect types present in an anomalous product. Additionally, our approach outperforms zero/few-shot learning SoTA methods on image-level and pixel-level anomaly detection and segmentation tasks on five commonly used datasets: MVTec-AD, Visa, MPDD, MAD and Real-IAD.
Related Topics
Concepts
No concepts available.
Metadata
- Type
- preprint
- Language
- en
- Landing Page
- http://arxiv.org/abs/2504.06740
- https://arxiv.org/pdf/2504.06740
- OA Status
- green
- OpenAlex ID
- https://openalex.org/W4417246815
All OpenAlex metadata
Raw OpenAlex JSON
- OpenAlex ID
-
https://openalex.org/W4417246815Canonical identifier for this work in OpenAlex
- DOI
-
https://doi.org/10.48550/arxiv.2504.06740Digital Object Identifier
- Title
-
MultiADS: Defect-aware Supervision for Multi-type Anomaly Detection and Segmentation in Zero-Shot LearningWork title
- Type
-
preprintOpenAlex work type
- Language
-
enPrimary language
- Publication year
-
2025Year of publication
- Publication date
-
2025-04-09Full publication date if available
- Authors
-
Ylli Sadikaj, Hongkuan Zhou, Lavdim Halilaj, Stefan Schmid, Steffen Staab, Claudia PlantList of authors in order
- Landing page
-
https://arxiv.org/abs/2504.06740Publisher landing page
- PDF URL
-
https://arxiv.org/pdf/2504.06740Direct link to full text PDF
- Open access
-
YesWhether a free full text is available
- OA status
-
greenOpen access status per OpenAlex
- OA URL
-
https://arxiv.org/pdf/2504.06740Direct OA link when available
- Cited by
-
0Total citation count in OpenAlex
Full payload
| id | https://openalex.org/W4417246815 |
|---|---|
| doi | https://doi.org/10.48550/arxiv.2504.06740 |
| ids.doi | https://doi.org/10.48550/arxiv.2504.06740 |
| ids.openalex | https://openalex.org/W4417246815 |
| fwci | |
| type | preprint |
| title | MultiADS: Defect-aware Supervision for Multi-type Anomaly Detection and Segmentation in Zero-Shot Learning |
| biblio.issue | |
| biblio.volume | |
| biblio.last_page | |
| biblio.first_page | |
| is_xpac | False |
| apc_list | |
| apc_paid | |
| language | en |
| locations[0].id | pmh:oai:arXiv.org:2504.06740 |
| 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 | cc-by-nc-sa |
| locations[0].pdf_url | https://arxiv.org/pdf/2504.06740 |
| locations[0].version | submittedVersion |
| locations[0].raw_type | text |
| locations[0].license_id | https://openalex.org/licenses/cc-by-nc-sa |
| locations[0].is_accepted | False |
| locations[0].is_published | False |
| locations[0].raw_source_name | |
| locations[0].landing_page_url | http://arxiv.org/abs/2504.06740 |
| locations[1].id | doi:10.48550/arxiv.2504.06740 |
| 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.2504.06740 |
| indexed_in | arxiv, datacite |
| authorships[0].author.id | https://openalex.org/A5026397263 |
| authorships[0].author.orcid | https://orcid.org/0000-0003-3739-4800 |
| authorships[0].author.display_name | Ylli Sadikaj |
| authorships[0].author_position | first |
| authorships[0].raw_author_name | Sadikaj, Ylli |
| authorships[0].is_corresponding | False |
| authorships[1].author.id | https://openalex.org/A5009003549 |
| authorships[1].author.orcid | https://orcid.org/0000-0001-8158-3081 |
| authorships[1].author.display_name | Hongkuan Zhou |
| authorships[1].author_position | middle |
| authorships[1].raw_author_name | Zhou, Hongkuan |
| authorships[1].is_corresponding | False |
| authorships[2].author.id | https://openalex.org/A5031926202 |
| authorships[2].author.orcid | |
| authorships[2].author.display_name | Lavdim Halilaj |
| authorships[2].author_position | middle |
| authorships[2].raw_author_name | Halilaj, Lavdim |
| authorships[2].is_corresponding | False |
| authorships[3].author.id | https://openalex.org/A5066080641 |
| authorships[3].author.orcid | https://orcid.org/0000-0002-7798-1711 |
| authorships[3].author.display_name | Stefan Schmid |
| authorships[3].author_position | middle |
| authorships[3].raw_author_name | Schmid, Stefan |
| authorships[3].is_corresponding | False |
| authorships[4].author.id | https://openalex.org/A5062807811 |
| authorships[4].author.orcid | https://orcid.org/0000-0002-0780-4154 |
| authorships[4].author.display_name | Steffen Staab |
| authorships[4].author_position | middle |
| authorships[4].raw_author_name | Staab, Steffen |
| authorships[4].is_corresponding | False |
| authorships[5].author.id | https://openalex.org/A5009516958 |
| authorships[5].author.orcid | https://orcid.org/0000-0001-5274-8123 |
| authorships[5].author.display_name | Claudia Plant |
| authorships[5].author_position | last |
| authorships[5].raw_author_name | Plant, Claudia |
| authorships[5].is_corresponding | False |
| has_content.pdf | True |
| has_content.grobid_xml | False |
| is_paratext | False |
| open_access.is_oa | True |
| open_access.oa_url | https://arxiv.org/pdf/2504.06740 |
| open_access.oa_status | green |
| open_access.any_repository_has_fulltext | False |
| created_date | 2025-10-10T00:00:00 |
| display_name | MultiADS: Defect-aware Supervision for Multi-type Anomaly Detection and Segmentation in Zero-Shot Learning |
| has_fulltext | False |
| is_retracted | False |
| updated_date | 2025-12-11T17:29:49.651110 |
| primary_topic | |
| cited_by_count | 0 |
| locations_count | 2 |
| best_oa_location.id | pmh:oai:arXiv.org:2504.06740 |
| 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 | cc-by-nc-sa |
| best_oa_location.pdf_url | https://arxiv.org/pdf/2504.06740 |
| best_oa_location.version | submittedVersion |
| best_oa_location.raw_type | text |
| best_oa_location.license_id | https://openalex.org/licenses/cc-by-nc-sa |
| 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/2504.06740 |
| primary_location.id | pmh:oai:arXiv.org:2504.06740 |
| 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 | cc-by-nc-sa |
| primary_location.pdf_url | https://arxiv.org/pdf/2504.06740 |
| primary_location.version | submittedVersion |
| primary_location.raw_type | text |
| primary_location.license_id | https://openalex.org/licenses/cc-by-nc-sa |
| primary_location.is_accepted | False |
| primary_location.is_published | False |
| primary_location.raw_source_name | |
| primary_location.landing_page_url | http://arxiv.org/abs/2504.06740 |
| publication_date | 2025-04-09 |
| publication_year | 2025 |
| referenced_works_count | 0 |
| abstract_inverted_index.a | 21, 39, 70, 93, 123, 141 |
| abstract_inverted_index.To | 127 |
| abstract_inverted_index.We | 90 |
| abstract_inverted_index.an | 181 |
| abstract_inverted_index.as | 38 |
| abstract_inverted_index.i) | 156 |
| abstract_inverted_index.if | 20 |
| abstract_inverted_index.in | 3, 58, 122, 146, 180 |
| abstract_inverted_index.is | 6, 23, 28, 72, 135 |
| abstract_inverted_index.it | 27 |
| abstract_inverted_index.of | 35, 55, 107, 130 |
| abstract_inverted_index.on | 80, 192, 201 |
| abstract_inverted_index.or | 25, 42, 74 |
| abstract_inverted_index.to | 30, 46, 66, 98, 115, 139, 150, 168 |
| abstract_inverted_index.MAD | 209 |
| abstract_inverted_index.The | 44, 105 |
| abstract_inverted_index.and | 12, 86, 103, 111, 119, 172, 194, 198, 210 |
| abstract_inverted_index.any | 78 |
| abstract_inverted_index.are | 64 |
| abstract_inverted_index.for | 8, 161 |
| abstract_inverted_index.ii) | 166 |
| abstract_inverted_index.not | 75 |
| abstract_inverted_index.our | 131, 133, 154, 185 |
| abstract_inverted_index.the | 14, 32, 48, 56, 81, 117, 128, 136, 151 |
| abstract_inverted_index.CLIP | 110 |
| abstract_inverted_index.SoTA | 190 |
| abstract_inverted_index.able | 97 |
| abstract_inverted_index.best | 129 |
| abstract_inverted_index.cut, | 41 |
| abstract_inverted_index.each | 162 |
| abstract_inverted_index.five | 202 |
| abstract_inverted_index.iii) | 173 |
| abstract_inverted_index.know | 31 |
| abstract_inverted_index.not, | 26 |
| abstract_inverted_index.such | 37 |
| abstract_inverted_index.task | 145 |
| abstract_inverted_index.type | 34, 51 |
| abstract_inverted_index.used | 204 |
| abstract_inverted_index.MPDD, | 208 |
| abstract_inverted_index.Visa, | 207 |
| abstract_inverted_index.align | 116 |
| abstract_inverted_index.bent, | 40 |
| abstract_inverted_index.extra | 112 |
| abstract_inverted_index.first | 137 |
| abstract_inverted_index.joint | 124 |
| abstract_inverted_index.masks | 160 |
| abstract_inverted_index.other | 152 |
| abstract_inverted_index.rates | 11 |
| abstract_inverted_index.scrap | 10 |
| abstract_inverted_index.tasks | 200 |
| abstract_inverted_index.type, | 83, 165 |
| abstract_inverted_index.types | 178 |
| abstract_inverted_index.costs. | 16 |
| abstract_inverted_index.defect | 50, 82, 164, 170, 177 |
| abstract_inverted_index.layers | 114 |
| abstract_inverted_index.learns | 167 |
| abstract_inverted_index.linear | 113 |
| abstract_inverted_index.lines. | 61 |
| abstract_inverted_index.merely | 18 |
| abstract_inverted_index.modern | 59 |
| abstract_inverted_index.solely | 67 |
| abstract_inverted_index.space. | 126 |
| abstract_inverted_index.types, | 171 |
| abstract_inverted_index."exact" | 49 |
| abstract_inverted_index.Anomaly | 101 |
| abstract_inverted_index.Besides | 17 |
| abstract_inverted_index.Current | 62 |
| abstract_inverted_index.Precise | 0 |
| abstract_inverted_index.ability | 45 |
| abstract_inverted_index.anomaly | 143, 159, 196 |
| abstract_inverted_index.crucial | 7, 29 |
| abstract_inverted_index.defect, | 36 |
| abstract_inverted_index.enables | 52 |
| abstract_inverted_index.feature | 125 |
| abstract_inverted_index.limited | 65 |
| abstract_inverted_index.methods | 63, 191 |
| abstract_inverted_index.optical | 1 |
| abstract_inverted_index.perform | 99, 140 |
| abstract_inverted_index.present | 179 |
| abstract_inverted_index.product | 22, 71 |
| abstract_inverted_index.propose | 91 |
| abstract_inverted_index.textual | 120 |
| abstract_inverted_index.visual- | 118 |
| abstract_inverted_index.whether | 69 |
| abstract_inverted_index.without | 76 |
| abstract_inverted_index.Contrary | 149 |
| abstract_inverted_index.MultiADS | 108 |
| abstract_inverted_index.approach | 138, 155, 186 |
| abstract_inverted_index.commonly | 203 |
| abstract_inverted_index.defects. | 89 |
| abstract_inverted_index.distinct | 33, 163 |
| abstract_inverted_index.insights | 79 |
| abstract_inverted_index.learning | 95, 189 |
| abstract_inverted_index.multiple | 88, 176 |
| abstract_inverted_index.product. | 183 |
| abstract_inverted_index.reducing | 13 |
| abstract_inverted_index.scratch. | 43 |
| abstract_inverted_index.specific | 158 |
| abstract_inverted_index.Detection | 102 |
| abstract_inverted_index.MVTec-AD, | 206 |
| abstract_inverted_index.MultiADS, | 92 |
| abstract_inverted_index.Real-IAD. | 211 |
| abstract_inverted_index.anomalies | 57 |
| abstract_inverted_index.anomalous | 24, 182 |
| abstract_inverted_index.approach, | 96 |
| abstract_inverted_index.automated | 53 |
| abstract_inverted_index.comprises | 109 |
| abstract_inverted_index.datasets: | 205 |
| abstract_inverted_index.defective | 73 |
| abstract_inverted_index.detecting | 19, 68, 85 |
| abstract_inverted_index.detection | 197 |
| abstract_inverted_index.generates | 157 |
| abstract_inverted_index.learning. | 148 |
| abstract_inverted_index.proposal, | 134 |
| abstract_inverted_index.providing | 77 |
| abstract_inverted_index.recognize | 47 |
| abstract_inverted_index.zero-shot | 94, 147 |
| abstract_inverted_index.Multi-type | 100 |
| abstract_inverted_index.associated | 15 |
| abstract_inverted_index.baselines, | 153 |
| abstract_inverted_index.identifies | 175 |
| abstract_inverted_index.industrial | 4 |
| abstract_inverted_index.inspection | 2 |
| abstract_inverted_index.knowledge, | 132 |
| abstract_inverted_index.minimizing | 9 |
| abstract_inverted_index.multi-type | 142 |
| abstract_inverted_index.production | 60 |
| abstract_inverted_index.treatments | 54 |
| abstract_inverted_index.distinguish | 169 |
| abstract_inverted_index.identifying | 87 |
| abstract_inverted_index.image-level | 193 |
| abstract_inverted_index.outperforms | 187 |
| abstract_inverted_index.pixel-level | 195 |
| abstract_inverted_index.applications | 5 |
| abstract_inverted_index.architecture | 106 |
| abstract_inverted_index.nevertheless | 84 |
| abstract_inverted_index.segmentation | 144, 199 |
| abstract_inverted_index.Additionally, | 184 |
| abstract_inverted_index.Segmentation. | 104 |
| abstract_inverted_index.zero/few-shot | 188 |
| abstract_inverted_index.representation | 121 |
| abstract_inverted_index.simultaneously | 174 |
| cited_by_percentile_year | |
| countries_distinct_count | 0 |
| institutions_distinct_count | 6 |
| citation_normalized_percentile |