{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,12]],"date-time":"2026-08-12T11:41:01Z","timestamp":1786534861874,"version":"3.56.0"},"reference-count":47,"publisher":"Wiley","license":[{"start":{"date-parts":[[2026,8,12]],"date-time":"2026-08-12T00:00:00Z","timestamp":1786492800000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"},{"start":{"date-parts":[[2026,8,12]],"date-time":"2026-08-12T00:00:00Z","timestamp":1786492800000},"content-version":"tdm","delay-in-days":0,"URL":"http:\/\/doi.wiley.com\/10.1002\/tdm_license_1.1"}],"funder":[{"DOI":"10.13039\/501100001821","name":"Vienna Science and Technology Fund","doi-asserted-by":"publisher","award":["1047379\/ICT22028"],"award-info":[{"award-number":["1047379\/ICT22028"]}],"id":[{"id":"10.13039\/501100001821","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["onlinelibrary.wiley.com"],"crossmark-restriction":true},"short-container-title":["Computer Graphics Forum"],"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>Computer vision increasingly uses synthetic data from physically based rendering to supplement limited real\u2010world datasets. In industrial inspection, defect data is scarce and the rendering pipeline is explicitly controlled, and synthetic defect generation therefore becomes a dataset design problem. In this setting, building a dataset means choosing points in a rendering parameter space: defect shape, material, illumination, viewpoint, and sampling define the training distribution, but their effect on downstream learning is often hard to judge from images alone. We therefore study how these factors change defect features, where their relation to downstream learning is easier to inspect. Our results show that the rendering factors do not matter equally: defect shape, material, illumination, and viewpoint often affect downstream behavior much more than the number of samples per pixel. Synthetic subsets that transfer better downstream tend to stay close to real defect features, cover the observed defect modes, and stay separated from the defect\u2010free (OK) region. Based on these observations, we build a simple feature\u2010space screening heuristic for selecting subsets from large candidate pools. The selected subsets often outperform matched random selection for downstream segmentation on real data.<\/jats:p>","DOI":"10.1111\/cgf.70550","type":"journal-article","created":{"date-parts":[[2026,8,12]],"date-time":"2026-08-12T11:18:11Z","timestamp":1786533491000},"update-policy":"https:\/\/doi.org\/10.1002\/crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Rendering Synthetic Defects for Learning\u2010Based Industrial Inspection"],"prefix":"10.1111","author":[{"ORCID":"https:\/\/orcid.org\/0009-0009-5230-4157","authenticated-orcid":false,"given":"Runzhou","family":"Mao","sequence":"first","affiliation":[{"name":"Fraunhofer ITWM  Germany"},{"name":"RPTU Kaiserslautern  Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-0388-8458","authenticated-orcid":false,"given":"Hiroyuki","family":"Sakai","sequence":"additional","affiliation":[{"name":"TU Wien  Austria"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4224-4105","authenticated-orcid":false,"given":"Christian","family":"Freude","sequence":"additional","affiliation":[{"name":"TU Wien  Austria"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-1669-8549","authenticated-orcid":false,"given":"Christoph","family":"Garth","sequence":"additional","affiliation":[{"name":"RPTU Kaiserslautern  Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-2102-2618","authenticated-orcid":false,"given":"Petra","family":"Gospodneti\u0107","sequence":"additional","affiliation":[{"name":"Fraunhofer ITWM  Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-1858-5978","authenticated-orcid":false,"given":"Juraj","family":"Fulir","sequence":"additional","affiliation":[{"name":"Fraunhofer ITWM  Germany"},{"name":"RPTU Kaiserslautern  Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"311","published-online":{"date-parts":[[2026,8,12]]},"reference":[{"key":"e_1_2_8_2_2","unstructured":"AmirS. GandelsmanY. BagonS. DekelT.: Deep ViT features as dense visual descriptors.CoRR abs\/2112.05814(122021). doi:10.48550\/arXiv.2112.05814. 4 5"},{"key":"e_1_2_8_3_2","unstructured":"Appleseed Team:Appleseed: A modern open source production renderer 2025. Accessed: 2026\u201004\u201012. URL:https:\/\/appleseedhq.net\/. 5"},{"key":"e_1_2_8_4_2","unstructured":"AbbasA. RusakE. TirumalaK. BrendelW. ChaudhuriK. MorcosA. S.: Effective pruning of Web\u2010Scale datasets based on complexity of concept clusters. InThe Twelfth International Conference on Learning Representations(2024). doi:10.48550\/arXiv.2401.04578. 3"},{"key":"e_1_2_8_5_2","doi-asserted-by":"crossref","unstructured":"BergmannP. FauserM. SattleggerD. StegerC.: MVTec AD \u2014 a comprehensive real\u2010world dataset for unsupervised anomaly detection.Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition(62019) 9592\u20139600. doi:10.1109\/cvpr.2019.00982. 1","DOI":"10.1109\/CVPR.2019.00982"},{"key":"e_1_2_8_6_2","doi-asserted-by":"publisher","DOI":"10.1007\/s11263\u2010020\u201001400\u20104"},{"key":"e_1_2_8_7_2","doi-asserted-by":"publisher","DOI":"10.1109\/MCG.2023.3243276"},{"key":"e_1_2_8_8_2","doi-asserted-by":"publisher","DOI":"10.1162\/neco.1995.7.1.108"},{"key":"e_1_2_8_9_2","doi-asserted-by":"publisher","DOI":"10.1111\/j.1467\u20108659.2004.00767.x"},{"key":"e_1_2_8_10_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.compind.2021.103459"},{"key":"e_1_2_8_11_2","doi-asserted-by":"publisher","DOI":"10.1145\/2343483.2343493"},{"key":"e_1_2_8_12_2","doi-asserted-by":"crossref","unstructured":"CaronM. TouvronH. MisraI. J\u00e9gouH. MairalJ. BojanowskiP. JoulinA.: Emerging properties in self\u2010supervised vision transformers. InProceedings of the IEEE\/CVF International Conference on Computer Vision(2021) pp.9650\u20139660. doi:10.1109\/ICCV48922.2021.00951. 4 5","DOI":"10.1109\/ICCV48922.2021.00951"},{"key":"e_1_2_8_13_2","doi-asserted-by":"publisher","DOI":"10.1038\/s44387-025-00024-8"},{"key":"e_1_2_8_14_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPRW59228.2023.00465"},{"key":"e_1_2_8_15_2","doi-asserted-by":"publisher","DOI":"10.3390\/s25196016"},{"key":"e_1_2_8_16_2","doi-asserted-by":"publisher","DOI":"10.3390\/s23135922"},{"key":"e_1_2_8_17_2","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1706.08500"},{"key":"e_1_2_8_18_2","doi-asserted-by":"publisher","DOI":"10.1109\/ICIP.2019.8803821"},{"key":"e_1_2_8_19_2","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1412.6980"},{"key":"e_1_2_8_20_2","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1904.06991"},{"key":"e_1_2_8_21_2","first-page":"281","volume-title":"Proceedings of the Fifth Berkeley Symposium on Mathematical Statistics and Probability, Volume 1: Statistics","author":"MacQueen J.","year":"1967"},{"key":"e_1_2_8_22_2","doi-asserted-by":"crossref","unstructured":"MaoR. FulirJ. GarthC. Gospodneti\u0107P.: Sequential PatchCore: Anomaly detection for surface inspection using synthetic impurities.Lecture Notes in Computer Science 15626 LNCS(2025) 360\u2013377. doi:10.1007\/978\u20103\u2010031\u201092805\u20105_23. 2 3","DOI":"10.1007\/978-3-031-92805-5_23"},{"key":"e_1_2_8_23_2","doi-asserted-by":"publisher","DOI":"10.2312\/sr.20251188"},{"key":"e_1_2_8_24_2","doi-asserted-by":"publisher","DOI":"10.1109\/ISIE45552.2021.9576231"},{"key":"e_1_2_8_25_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.promfg.2017.07.198"},{"key":"e_1_2_8_26_2","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2002.09797"},{"key":"e_1_2_8_27_2","unstructured":"OquabM. DarcetT. MoutakanniT. VoH. V. SzafraniecM. KhalidovV. FernandezP. HazizaD. MassaF. El\u2010NoubyA. AssranM. BallasN. GalubaW. HowesR. HuangP. Y. LiS. W. MisraI. RabbatM. SharmaV. SynnaeveG. XuH. JegouH. MairalJ. LabatutP. JoulinA. BojanowskiP.: DINOv2: Learning robust visual features without supervision.Transactions on Machine Learning Research 2024(42023). doi:10.48550\/arXiv.2304.07193. 4 5"},{"key":"e_1_2_8_28_2","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2107.07075"},{"key":"e_1_2_8_29_2","volume-title":"Physically Based Rendering: From Theory to Implementation","author":"Pharr M.","year":"2016"},{"key":"e_1_2_8_30_2","doi-asserted-by":"publisher","DOI":"10.1016\/0377\u20100427(87)90125\u20107"},{"key":"e_1_2_8_31_2","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr52688.2022.01392"},{"key":"e_1_2_8_32_2","doi-asserted-by":"publisher","DOI":"10.1109\/wacv56688.2023.00262"},{"key":"e_1_2_8_33_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.procir.2022.05.115"},{"key":"e_1_2_8_34_2","doi-asserted-by":"publisher","DOI":"10.52202\/068431-1419"},{"key":"e_1_2_8_35_2","doi-asserted-by":"publisher","DOI":"10.1016\/J.COMPIND.2025.104360"},{"key":"e_1_2_8_36_2","doi-asserted-by":"publisher","DOI":"10.3390\/s24020649"},{"key":"e_1_2_8_37_2","unstructured":"SenerO. SavareseS.: Active learning for convolutional neural networks: A core\u2010set approach. InInternational Conference on Learning Representations(2018). doi:10.48550\/arXiv.1708.00489. 3"},{"key":"e_1_2_8_38_2","doi-asserted-by":"crossref","unstructured":"TobinJ. FongR. RayA. SchneiderJ. ZarembaW. AbbeelP.: Domain randomization for transferring deep neural networks from simulation to the real world.IEEE International Conference on Intelligent Robots and Systems 2017\u2010September(32017) 23\u201330. doi:10.1109\/iros.2017.8202133. 2 3 8","DOI":"10.1109\/IROS.2017.8202133"},{"key":"e_1_2_8_39_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPRW.2018.00143"},{"key":"e_1_2_8_40_2","volume-title":"Information Retrieval","author":"Rijsbergen C. J.","year":"1979"},{"key":"e_1_2_8_41_2","doi-asserted-by":"publisher","DOI":"10.1109\/TIP.2003.819861"},{"key":"e_1_2_8_42_2","doi-asserted-by":"publisher","DOI":"10.1109\/ETFA61755.2024.10710816"},{"key":"e_1_2_8_43_2","unstructured":"WangR. HoppeS. MonariE. HuberM. F.: Defect Transfer GAN: Diverse defect synthesis for data augmentation. In33rd British Machine Vision Conference (BMVC 2022)(2022) BMVA Press. doi:10.48550\/arXiv.2302.08366. 2 3"},{"key":"e_1_2_8_44_2","unstructured":"WalterB. MarschnerS. R. LiH. TorranceK. E.: Microfacet models for refraction through rough surfaces.Eurographics Symposium on Rendering(2007) 195\u2013206. doi:10.2312\/EGWR\/EGSR07\/195\u2010206. 7"},{"key":"e_1_2_8_45_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPRW59228.2023.00467"},{"key":"e_1_2_8_46_2","doi-asserted-by":"publisher","DOI":"10.1109\/WACV48630.2021.00257"},{"key":"e_1_2_8_47_2","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr.2018.00068"},{"key":"e_1_2_8_48_2","unstructured":"ZhengH. LiuR. LaiF. PrakashA.: Coverage\u2010centric coreset selection for high pruning rates. InInternational Conference on Learning Representations(2023). doi:10.48550\/arXiv.2210.15809. 3"}],"container-title":["Computer Graphics Forum"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/onlinelibrary.wiley.com\/doi\/pdf\/10.1111\/cgf.70550","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/onlinelibrary.wiley.com\/doi\/full-xml\/10.1111\/cgf.70550","content-type":"application\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/onlinelibrary.wiley.com\/doi\/pdf\/10.1111\/cgf.70550","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,12]],"date-time":"2026-08-12T11:18:21Z","timestamp":1786533501000},"score":1,"resource":{"primary":{"URL":"https:\/\/onlinelibrary.wiley.com\/doi\/10.1111\/cgf.70550"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,8,12]]},"references-count":47,"alternative-id":["10.1111\/cgf.70550"],"URL":"https:\/\/doi.org\/10.1111\/cgf.70550","archive":["Portico"],"relation":{},"ISSN":["0167-7055","1467-8659"],"issn-type":[{"value":"0167-7055","type":"print"},{"value":"1467-8659","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,8,12]]},"assertion":[{"value":"2026-08-12","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}],"article-number":"e70550"}}