Logo do repositório
 
Publicação

Unification of Closed-Open Industrial Detection Scenarios

2026-08Artigo científico Acesso embargado
Ver registo simples
dc.contributor.authorZhang, Zekai
dc.contributor.authorZhang, Jinglin
dc.contributor.authorChen, Qinghui
dc.contributor.authorLi, Gang
dc.contributor.authorChen, Da
dc.contributor.authorJing, Shuainan
dc.contributor.authorWang, He
dc.contributor.authorLi, Dagang
dc.contributor.authorLiu, Cong
dc.contributor.authorBai, Cong
dc.contributor.authorChen, Shengyong
dc.contributor.institutionNOVA Information Management School (NOVA IMS)
dc.contributor.institutionInformation Management Research Center (MagIC) - NOVA Information Management School
dc.contributor.pblIEEE Computer Society
dc.date.accessioned2026-07-13T10:51:02Z
dc.date.available2026-07-13T10:51:02Z
dc.date.embargoedUntil2028-04-03
dc.date.issued2026-08
dc.descriptionZhang, Z., Zhang, J., Chen, Q., Li, G., Chen, D., Jing, S., Wang, H., Li, D., Liu, C., Bai, C., & Chen, S. (2026). Unification of Closed-Open Industrial Detection Scenarios: New Large-Scale Benchmarks, Challenges and Baselines. IEEE Transactions on Pattern Analysis and Machine Intelligence, 8(48), 9571-9588. https://doi.org/10.1109/TPAMI.2026.3680569
dc.description.abstractLarge-scale Visual-Language Models (LVLMs) have achieved remarkable success in natural visual tasks, yet their application to industrial defect detection remains challenging due to two fundamental limitations: (i) the scarcity of large-scale industrial datasets that cover diverse defect categories across multiple domains, and (ii) the reliance on manual prompts (points, boxes, masks) that introduce subjective noise and lack text-visual interaction for fine-grained understanding. To address these challenges, we introduce a Large-Scale Multi-Modal Industrial Open-Closed benchmark (MMIOC-1M) containing over one million samples across 14 super-categories, 29 industrial scenes, and 351 defect subcategories. To our knowledge, MMIOC-1M is the first unified largest benchmark supporting both open-vocabulary and closed-set industrial detection, providing valuable pre-training data for LVLMs in industrial scenarios. Furthermore, we propose a Refined Text-Visual Prompt Network (RTVPNet) that incorporates three key innovations: (1) an expert-assisted domain projection mechanism that enables rapid adaptation of general vision models to industrial domains, (2) an energy-based sparse sampling strategy that automatically generates refined visual prompts without manual intervention, and (3) a bidirectional text-visual interaction module that enhances cross-modal semantic alignment and understanding. Extensive experiments demonstrate that RTVPNet achieves state-of-the-art performance on MMIOC-1M, LVIS, and COCO benchmarks while maintaining computational efficiency. The dataset and code are available at https://github.com/hellozzk/MMIO.en
dc.description.versionauthorsversion
dc.description.versionpublished
dc.format.extent18
dc.format.extent20443674
dc.identifier.doi10.1109/TPAMI.2026.3680569
dc.identifier.issn0162-8828
dc.identifier.otherPURE: 159075177
dc.identifier.otherPURE UUID: 4624640f-6ca4-4cbb-addd-218f4633ba6d
dc.identifier.otherScopus: 105034860840
dc.identifier.otherWOS: 001815311000033
dc.identifier.urihttp://hdl.handle.net/10362/204450
dc.identifier.urlhttps://www.scopus.com/pages/publications/105034860840
dc.identifier.urlhttps://www.webofscience.com/wos/woscc/full-record/WOS:001815311000033
dc.language.isoeng
dc.peerreviewedyes
dc.relationhttps://doi.org/10.54499/UID/04152/2025
dc.relationhttps://doi.org/10.54499/UID/PRR/04152/2025
dc.subjectIndustrial Open Detection
dc.subjectLarge Scale Industrial Benchmark
dc.subjectVisual Language Model
dc.subjectSoftware
dc.subjectComputer Vision and Pattern Recognition
dc.subjectComputational Theory and Mathematics
dc.subjectApplied Mathematics
dc.subjectArtificial Intelligence
dc.subjectSDG 9 - Industry, Innovation, and Infrastructure
dc.titleUnification of Closed-Open Industrial Detection Scenariosen
dc.title.subtitleNew Large-Scale Benchmarks, Challenges and Baselinesen
dc.typejournal article
degois.publication.firstPage9571
degois.publication.issue48
degois.publication.lastPage9588
degois.publication.titleIEEE Transactions on Pattern Analysis and Machine Intelligence
degois.publication.volume8
dspace.entity.typePublication
rcaap.rightsembargoedAccess

Ficheiros

Principais
A mostrar 1 - 1 de 1
Miniatura indisponível
Nome:
Unification_of_Closed-Open_AAM.pdf
Tamanho:
19.5 MB
Formato:
Adobe Portable Document Format
Ver/Abrir

Coleções

NIMS: MagIC - Artigos em revista internacional com arbitragem científica (Peer-Review articles in international journals)