面向低样本的工业图像异常检测综述

doi:10.3778/j.issn.1002-8331.2408-0230

摘要/Abstract

摘要： 异常检测是计算机视觉的重要研究方向之一，被广泛应用于工业产品检测、医疗诊断和视频监控等领域。它可以监控产品质量并快速识别不符合标准的产品，从而实现自动化质量控制。在工业图像异常检测中，由于获取大量训练样本既耗时又昂贵，因此低样本的工业图像异常检测成为研究热点和趋势。同时，大型视觉语言模型的出现推动了工业图像异常检测从单模态驱动发展到多模态融合，提高了检测的全面性。针对近四年的低样本工业图像异常检测，探讨了模态数量对异常检测性能的影响，并根据检测算法在处理数据和检测异常时采用的不同策略以及特征提取技术的提取范围对低样本异常检测方法进行分类。这旨在帮助研究人员快速了解并进一步改进异常检测技术和优化特征提取策略，从而提升低样本环境下的检测效率和准确性。此外，还对不同方法在MVTec AD和VisA数据集上的检测结果进行了对比，对比结果表明多模态融合方法中的混合特征融合在分类和分割异常方面表现突出。通过对比不同的检测技术和模型，总结了不同的低样本工业异常检测方法以及解决的问题，并讨论了未来的研究方向。

关键词: 工业图像异常检测, 低样本, 单模态驱动, 多模态融合, 混合特征融合

Abstract: Anomaly detection is a significant research direction in computer vision, widely applied in industrial product inspection, medical diagnostics, and video surveillance, among other fields. It can monitor product quality and swiftly identify products that do not meet standards, thereby achieving automated quality control. In industrial image anomaly detection, acquiring a large number of training samples is both time-consuming and costly, making low-shot industrial image anomaly detection a hot topic and trend. Meanwhile, the emergence of large-scale visual-language models has propelled industrial image anomaly detection from being driven by a single-modal driven to a multi-modal fusion, enhancing the comprehensiveness of detection. Focusing on the past four years of low-shot industrial image anomaly detection, this study explores the impact of modality quantity on anomaly detection performance, and classifies low-shot industrial image anomaly detection methods based on the different strategies adopted in handling data and detecting anomalies as well as the extraction range of feature extraction technologies. The aim is to help researchers quickly understand and further improve anomaly detection technologies and optimize feature extraction strategies, thereby enhancing detection efficiency and accuracy under low-shot conditions. Additionally, the detection results of different methods on the MVTec AD and VisA datasets are compared, showing that hybrid feature fusion in multi-modal fusion methods performs exceptionally well in anomaly classification and segmentation. By comparing different detection techniques and models, this study summarizes various low-shot industrial anomaly detection methods and the problems they address, and discusses future research directions.

Key words: industrial image anomaly detection, low-shot, single-modal driven, multi-modal fusion, hybrid feature fusion

郭新茹, 宋丽娟, 朱文倩, 杜方, 马子睿. 面向低样本的工业图像异常检测综述[J]. 计算机工程与应用, 2025, 61(13): 26-45.

GUO Xinru, SONG Lijuan, ZHU Wenqian, DU Fang, MA Zirui. Review of Low-Shot Industrial Image Anomaly Detection[J]. Computer Engineering and Applications, 2025, 61(13): 26-45.

参考文献

[1] XU H H, XU S C, YANG W Z. Unsupervised industrial anomaly detection with diffusion models[J]. Journal of Visual Communication and Image Representation, 2023, 97: 103983.
[2] MANIVANNAN S. Automatic quality inspection in additive manufacturing using semi-supervised deep learning[J]. Journal of Intelligent Manufacturing, 2023, 34(7): 3091-3108.
[3] XIE G Y, WANG J B, LIU J Q, et al. IM-IAD: industrial image anomaly detection benchmark in manufacturing[J]. IEEE Transactions on Cybernetics, 2024, 54(5): 2720-2733.
[4] QU X J, LIU Z, WU C Q, et al. MFGAN: multimodal fusion for industrial anomaly detection using attention-based autoencoder and generative adversarial network[J]. Sensors, 2024, 24(2): 637.
[5] LI Y W, IVANOVA E, BRUVERIS M. FADE: few-shot/zero-shot anomaly detection engine using large vision-language model[J]. arXiv:2409.00556, 2024.
[6] LI H X, WU J Q, WU L Y, et al. Towards efficient pixel labeling for industrial anomaly detection and localization[J]. arXiv:2407.03130, 2024.
[7] YANG J, SHI Y, QI Z Q. Learning deep feature correspondence for unsupervised anomaly detection and segmentation[J]. Pattern Recognition, 2022, 132: 108874.
[8] HE K M, ZHANG X Y, REN S Q, et al. Deep residual learning for image recognition[C]//Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2016: 770-778.
[9] KINGMA D P, WELLING M. Auto-encoding variational Bayes[J]. arXiv:1312.6114, 2013.
[10] DOERSCH C. Tutorial on variational autoencoders[J]. arXiv:1606.05908, 2016.
[11] GOODFELLOW I J, POUGET-ABADIE J, MIRZA M, et al. Generative adversarial networks[C]//Advances in Neural Information Processing Systems, 2014.
[12] VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need[C]//Advances in Neural Information Processing Systems, 2017.
[13] DOSOVITSKIY A, BEYER L, KOLESNIKOV A, et al. An image is worth 16x16 words: transformers for image recognition at scale[J]. arXiv:2010.11929, 2020.
[14] 孙露露, 刘建平, 王健, 等. 细粒度图像分类上Vision Transformer的发展综述[J]. 计算机工程与应用, 2024, 60(10): 30-46.
SUN L L, LIU J P, WANG J, et al. Survey of vision transformer in fine-grained image classification[J]. Computer Engineering and Applications, 2024, 60(10): 30-46.
[15] RADFORD A, KIM J W, HALLACY C, et al. Learning transferable visual models from natural language supervision[C]//Proceedings of the International Conference on Machine Learning, 2021: 8748-8763.
[16] KIRILLOV A, MINTUN E, RAVI N, et al. Segment anything[C]//Proceedings of the 2023 IEEE/CVF International Conference on Computer Vision. Piscataway: IEEE, 2023: 3992-4003.
[17] GUO Y H, SONG L J, MA Z R. Self-supervised texture image anomaly detection by fusing normalizing flow and dictionary learning[J]. International Journal of Pattern Recognition and Artificial Intelligence, 2023, 37(4): 2351007.
[18] WANG J P, XU G L, LI C L, et al. Multi-feature reconstruction network using crossed-mask restoration for unsupervised industrial anomaly detection[J]. arXiv:2404.13273, 2024.
[19] ZHAN J W, LAI J X, GAO B B, et al. Enhancing multi-class anomaly detection via diffusion refinement with dual conditioning[J]. arXiv:2407.01905, 2024.
[20] TAO X, GONG X Y, ZHANG X, et al. Deep learning for unsupervised anomaly localization in industrial images: a survey[J]. IEEE Transactions on Instrumentation and Measurement, 2022, 71: 5018021.
[21] DIERS J, PIGORSCH C. A survey of methods for automated quality control based on images[J]. International Journal of Computer Vision, 2023, 131(10): 2553-2581.
[22] CUI Y J, LIU Z X, LIAN S G. A survey on unsupervised anomaly detection algorithms for industrial images[J]. IEEE Access, 2023, 11: 55297-55315.
[23] LIU J Q, XIE G Y, WANG J B, et al. Deep industrial image anomaly detection: a survey[J]. Machine Intelligence Research, 2024, 21(1): 104-135.
[24] 罗东亮, 蔡雨萱, 杨子豪, 等. 工业缺陷检测深度学习方法综述[J]. 中国科学: 信息科学, 2022, 52(6): 1002-1039.
LUO D L, CAI Y X, YANG Z H, et al. Survey on industrial defect detection with deep learning[J]. Scientia Sinica (Informationis), 2022, 52(6): 1002-1039.
[25] 周茂袁, 伍小双. 基于深度学习的异常检测模型综述[J]. 中国民航大学学报, 2023, 41(4): 1-7.
ZHOU M Y, WU X S. Review of anomaly detection models based on deep learning[J]. Journal of Civil Aviation University of China, 2023, 41(4): 1-7.
[26] 张邹铨, 张辉, 吴天月, 等. 面向工业图像异常检测的连续密集标准化流模型[J]. 计算机科学, 2023, 50(12): 212-220.
ZHANG Z Q, ZHANG H, WU T Y, et al. Continuous dense normalized flow model for anomaly detection in industrial images[J]. Computer Science, 2023, 50(12): 212-220.
[27] RUDOLPH M, WANDT B, ROSENHAHN B. Same same but DifferNet: semi-supervised defect detection with normalizing flows[C]//Proceedings of the 2021 IEEE Winter Conference on Applications of Computer Vision. Piscataway: IEEE, 2021: 1906-1915.
[28] DINH L, SOHL-DICKSTEIN J, BENGIO S. Density estimation using real NVP[J]. arXiv:1605.08803, 2016.
[29] LI X, LI S. Fast and few-shot unsupervised defect detection on flexible packaging surfaces[J]. IEEE Journal of Radio Frequency Identification, 2022, 6: 842-845.
[30] JADERBERG M, SIMONYAN K, ZISSERMAN A. Spatial transformer networks[C]//Advances in Neural Information Processing Systems, 2015.
[31] LIAO J Y, XU X, NGUYEN M C, et al. COFT-AD: contrastive fine-tuning for few-shot anomaly detection[J]. IEEE Transactions on Image Processing, 2024, 33: 2090-2103.
[32] BAI Y H, ZHANG J N, CHEN Z F, et al. Dual-path frequency discriminators for few-shot anomaly detection[J]. arXiv:2403. 04151, 2024.
[33] KAMOONA A M, GOSTAR A K, WANG X Y, et al. Anomaly detection of defect using energy of point pattern features within random finite set framework[J]. Engineering Applications of Artificial Intelligence, 2024, 130: 107706.
[34] DUSMANU M, ROCCO I, PAJDLA T, et al. D2-net: a trainable CNN for joint description and detection of local features[C]//Proceedings of the 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2019: 8084-8093.
[35] LI A, QIU C, KLOFT M, et al. Zero-shot anomaly detection via batch normalization[C]//Advances in Neural Information Processing Systems, 2024 .
[36] NAQI M M, ATIF TAHIR M. Industrial anomaly detection using vision based cross transformers[C]//Proceedings of the 2023 International Conference on IT and Industrial Technologies. Piscataway: IEEE, 2023: 1-5.
[37] HE A F, LUO C, TIAN X M, et al. A twofold Siamese network for real-time object tracking[C]//Proceedings of the 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2018: 4834-4843.
[38] HUANG C Q, GUAN H Y, JIANG A F, et al. Registration based few-shot anomaly detection[C]//Proceedings of the European Conference on Computer Vision. Cham: Springer, 2022: 303-319.
[39] SCHWARTZ E, ARBELLE A, KARLINSKY L, et al. MAEDAY: MAE for few- and zero-shot AnomalY-Detection[J]. Computer Vision and Image Understanding, 2024, 241: 103958.
[40] HE K M, CHEN X L, XIE S N, et al. Masked autoencoders are scalable vision learners[C]//Proceedings of the 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2022: 15979-15988.
[41] HU E J, SHEN Y, WALLIS P, et al. LoRA: low-rank adaptation of large language models[J]. arXiv:2106.09685, 2021.
[42] YE S J, WANG Z, XIONG P B, et al. Multi-stage few-shot micro-defect detection of patterned OLED panel using defect inpainting and multi-scale Siamese neural network[J]. Journal of Intelligent Manufacturing, 2024, 35(6): 2653-2669.
[43] WANG Z, BOVIK A C, SHEIKH H R, et al. Image quality assessment: from error visibility to structural similarity[J]. IEEE Transactions on Image Processing, 2004, 13(4): 600-612.
[44] 张玥, 陈锡伟, 陈梦丹, 等. 基于对比学习生成对抗网络的无监督工业品表面异常检测[J]. 电子测量与仪器学报, 2023, 37(10): 193-201.
ZHANG Y, CHEN X W, CHEN M D, et al. Unsupervised surface anomaly detection of industrial products based on contrastive learning generative adversarial network[J]. Journal of Electronic Measurement and Instrumentation, 2023, 37(10): 193-201.
[45] GONG Y, LIU M Z, WANG X Q, et al. Few-shot defect detection using feature enhancement and image generation for manufacturing quality inspection[J]. Applied Intelligence, 2024, 54(1): 375-397.
[46] SHEYNIN S, BENAIM S, WOLF L. A hierarchical transformation-discriminating generative model for few shot anomaly detection[C]//Proceedings of the 2021 IEEE/CVF International Conference on Computer Vision. Piscataway: IEEE, 2021: 8475-8484.
[47] LEE J Y, LEE W, CHOI J, et al. Few-shot anomaly detection with adversarial loss for robust feature representations[J]. arXiv:2312.03005, 2023.
[48] FANG Z, WANG X Y, LI H C, et al. FastRecon: few-shot industrial anomaly detection via fast feature reconstruction[C]//Proceedings of the 2023 IEEE/CVF International Conference on Computer Vision. Piscataway: IEEE, 2023: 17435-17444.
[49] YAO H M, LUO W, CAO Y K, et al. Global-regularized neighborhood regression for efficient zero-shot texture anomaly detection[J]. arXiv:2406.07333, 2024.
[50] 刘永江, 陈斌. 基于多尺度记忆库的像素级无监督工业异常检测[J]. 计算机应用, 2024, 44(11): 3587-3594.
LIU Y J, CHEN B. Pixel-level unsupervised industrial anomaly detection based on multi-scale memory bank[J]. Journal of Computer Applications, 2024, 44(11): 3587-3594.
[51] PENG Y, LIN X, MA N C, et al. SAM-LAD: segment anything model meets zero-shot logic anomaly detection[J]. arXiv:2406.00625, 2024.
[52] ROTH K, PEMULA L, ZEPEDA J, et al. Towards total recall in industrial anomaly detection[C]//Proceedings of the 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2022: 14298-14308.
[53] SANTOS J, TRAN T, RIPPEL O. Optimizing PatchCore for few/many-shot anomaly detection[J]. arXiv:2307.10792, 2023.
[54] XIE G Y, WANG J B, LIU J Q, et al. Pushing the limits of fewshot anomaly detection in industry vision: graphcore[J]. arXiv:2301.12082, 2023.
[55] AOTA T, TONG L T T, OKATANI T. Zero-shot versus many-shot: unsupervised texture anomaly detection[C]//Proceedings of the 2023 IEEE/CVF Winter Conference on Applications of Computer Vision. Piscataway: IEEE, 2023: 5553-5561.
[56] COHEN N, HOSHEN Y. Sub-image anomaly detection with deep pyramid correspondences[J]. arXiv:2005.02357, 2020.
[57] ZHANG L R, ZHANG S H, XIE G Y, et al. What makes a good data augmentation for few-shot unsupervised image anomaly detection?[C]//Proceedings of the 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops. Piscataway: IEEE, 2023: 4345-4354.
[58] DAMM S, LASZKIEWICZ M, LEDERER J, et al. Anomaly- DINO: boosting patch-based few-shot anomaly detection with DINOv2[J]. arXiv:2405.14529, 2024.
[59] MIN Y Z, WANG Z W, LIU Y, et al. FS-RSDD: few-shot rail surface defect detection with prototype learning[J]. Sensors, 2023, 23(18): 7894.
[60] 陈柳任. 基于知识蒸馏的工业图像异常检测算法研究[D]. 南京: 南京信息工程大学, 2023.
CHEN L R. Research on industrial image anomaly detection algorithm based on knowledge distillation[D]. Nanjing: Nanjing University of Information Science & Technology, 2023.
[61] GU Z H, YANG T H, MA L Z. Prototype-aware contrastive knowledge distillation for few-shot anomaly detection[C]//Proceedings of the 34th British Machine Vision Conference, 2023.
[62] DENG H Q, LI X Y. Anomaly detection via reverse distillation from one-class embedding[C]//Proceedings of the 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2022: 9727-9736.
[63] BELTON N, HAGOS M T, LAWLOR A, et al. FewSOME: one-class few shot anomaly detection with Siamese networks[C]//Proceedings of the 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops. Piscataway: IEEE, 2023: 2978-2987.
[64] WU J K, JIANG W J, HUANG Z Y, et al. Multi-scale prototypical network for few-shot anomaly detection[C]//Proceedings of the International Conference on Natural Computation, Fuzzy Systems and Knowledge Discovery. Cham: Springer, 2023: 1067-1076.
[65] SHI X W, ZHANG S B, CHENG M, et al. Few-shot semantic segmentation for industrial defect recognition[J]. Computers in Industry, 2023, 148: 103901.
[66] SHI X W, CUI Z, ZHANG S B, et al. Multi-similarity based hyperrelation network for few-shot segmentation[J]. IET Image Processing, 2023, 17(1): 204-214.
[67] LI X R, HUANG Z M, XUE F, et al. MuSc: zero-shot industrial anomaly classification and segmentation with mutual scoring of the unlabeled images[J]. arXiv:2401.16753, 2024.
[68] ZHANG J, SUGANUMA M, OKATANI T. Network pruning and fine-tuning for few-shot industrial image anomaly detection[C]//Proceedings of the 2023 IEEE 21st International Conference on Industrial Informatics. Piscataway: IEEE, 2023: 1-6.
[69] LI G, JAMPANI V, SEVILLA-LARA L, et al. Adaptive prototype learning and allocation for few-shot segmentation[C]//Proceedings of the 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2021: 8330-8339.
[70] ZHUANG Z S, ZHANG J H, XIAO L W, et al. PGTNet: prototype guided transfer network for few-shot anomaly localization[C]//Proceedings of the 2022 IEEE International Conference on Image Processing. Piscataway: IEEE, 2022: 2321-2325.
[71] 黄超. 工业图像无监督异常检测与定位研究[D]. 成都: 电子科技大学, 2023.
HUANG C. Unsupervised anomaly detection and localization in industrial images[D]. Chengdu: University of Electronic Science and Technology of China, 2023.
[72] BAO Y Q, SONG K C, LIU J, et al. Triplet-graph reasoning network for few-shot metal generic surface defect segmentation[J]. IEEE Transactions on Instrumentation and Measurement, 2021, 70: 5011111.
[73] ZHANG Z L, NIU C, ZHAO Z B, et al. Small object few-shot segmentation for vision-based industrial inspection[J]. arXiv:2407.21351, 2024.
[74] DEFARD T, SETKOV A, LOESCH A, et al. PaDiM: a patch distribution modeling framework for anomaly detection and localization[C]//Proceedings of the International Conference on Pattern Recognition. Cham: Springer, 2021: 475-489.
[75] WEI S X, WEI X, MA Z H, et al. Few-shot online anomaly detection and segmentation[J]. Knowledge-Based Systems, 2024, 300: 112168.
[76] ARDELEAN A T, WEYRICH T. High-fidelity zero-shot texture anomaly localization using feature correspondence analysis[C]//Proceedings of the 2024 IEEE/CVF Winter Conference on Applications of Computer Vision. Piscataway: IEEE, 2024: 1123-1133.
[77] TAMURA M. Random word data augmentation with CLIP for zero-shot anomaly detection[J]. arXiv:2308.11119, 2023.
[78] CAO Y K, XU X H, SUN C, et al. Towards generic anomaly detection and understanding: large-scale visual-linguistic model (GPT-4V) takes the lead[J]. arXiv:2311.02782, 2023.
[79] CHEN X H, ZHANG J N, TIAN G Z, et al. CLIP-AD: a language-guided staged dual-path model for zero-shot anomaly detection[J]. arXiv:2311.00453, 2023.
[80] LI Y, WANG H, DUAN Y, et al. Clip surgery for better explainability with enhancement in open-vocabulary tasks[J]. arXiv:2304.05653, 2023.
[81] KWAK S, JEONG J, LEE H, et al. Few-shot anomaly detection via personalization[J]. IEEE Access, 2024, 12: 11035-11051.
[82] ROMBACH R, BLATTMANN A, LORENZ D, et al. High-resolution image synthesis with latent diffusion models[C]//Proceedings of the 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2022: 10674-10685.
[83] GU Z P, ZHU B K, ZHU G B, et al. AnomalyGPT: detecting industrial anomalies using large vision-language models[J]. Proceedings of the AAAI Conference on Artificial Intelligence, 2024, 38(3): 1932-1940.
[84] JEONG J, ZOU Y, KIM T, et al. WinCLIP: zero-/ few-shot anomaly classification and segmentation[C]//Proceedings of the 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2023: 19606-19616.
[85] CHEN X H, HAN Y, ZHANG J N. APRIL-GAN: a zero-/ few-shot anomaly classification and segmentation method for CVPR 2023 VAND workshop challenge tracks 1&2: 1st place on zero-shot AD and 4th place on few-shot AD[J]. arXiv:2305.17382, 2023.
[86] DENG H Q, ZHANG Z X, BAO J N, et al. AnoVL: adapting vision-language models for unified zero-shot anomaly locali- zation[J]. arXiv:2308.15939, 2023.
[87] BAUGH M, BATTEN J, MüLLER J P, et al. Zero-shot anomaly detection with pre-trained segmentation models[J]. arXiv:2306.09269, 2023.
[88] QIN X B, DAI H, HU X B, et al. Highly accurate dichotomous image segmentation[C]//Proceedings of the European Conference on Computer Vision. Cham: Springer, 2022: 38-56.
[89] QU Z, TAO X, PRASAD M, et al. VCP-CLIP: a visual context prompting model for zero-shot anomaly segmentation[J]. arXiv:2407.12276, 2024.
[90] ZHU J W, PANG G S. Toward generalist anomaly detection via in-context residual learning with few-shot sample prompts[C]//Proceedings of the 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2024: 17826-17836.
[91] ZHOU Q H, PANG G S, TIAN Y, et al. AnomalyCLIP: object-agnostic prompt learning for zero-shot anomaly detection[J]. arXiv:2310.18961, 2023.
[92] HU Z X, ZHANG Z S. SOWA: adapting hierarchical frozen window self-attention to visual-language models for better anomaly detection[J]. arXiv:2407.03634, 2024.
[93] LI X F, ZHANG Z Z, TAN X, et al. PromptAD: learning prompts with only normal samples for few-shot anomaly detection[C]//Proceedings of the 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2024: 16848-16858.
[94] CAO Y K, ZHANG J N, FRITTOLI L, et al. AdaCLIP: adapting CLIP with hybrid learnable prompts for zero-shot anomaly detection[J]. arXiv:2407.5795, 2024.
[95] CAO Y K, XU X H, SUN C, et al. Segment any anomaly without training via hybrid prompt regularization[J]. arXiv: 2305.10724, 2023.
[96] LIU S L, ZENG Z Y, REN T H, et al. Grounding DINO: marrying DINO with grounded pre-training for open-set object detection[J]. arXiv:2303.05499, 2023.
[97] YAN Z Y, FANG Q Q, LV W X, et al. AnomalySD: few-shot multi-class anomaly detection with stable diffusion model[J]. arXiv:2408.01960, 2024.
[98] HUANG C Q, JIANG A F, ZHANG Y, et al. Multi-scale memory comparison for zero-/few-shot anomaly detection[J]. arXiv:2308.04789, 2023.
[99] LI S Z, CAO J J, YE P, et al. ClipSAM: CLIP and SAM collaboration for zero-shot anomaly segmentation[J]. arXiv: 2401.12665, 2024.
[100] QIN J J, GU C Z, YU J, et al. Teacher-student network for 3D point cloud anomaly detection with few normal samples[J]. Expert Systems with Applications, 2023, 228: 120371.
[101] WANG Y, PENG J L, ZHANG J N, et al. Multimodal industrial anomaly detection via hybrid fusion[C]//Proceedings of the 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2023: 8032-8041.
[102] COSTANZINO A, RAMIREZ P Z, LISANTI G, et al. Multimodal industrial anomaly detection by crossmodal feature mapping[C]//Proceedings of the 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2024: 17234-17243.
[103] ZUO Z, DONG J H, WU Y, et al. CLIP3D-AD: extending CLIP for 3D few-shot anomaly detection with multi-view images generation[J]. arXiv:2406.18941, 2024.
[104] BERGMANN P, BATZNER K, FAUSER M, et al. The MVTec anomaly detection dataset: a comprehensive real-world dataset for unsupervised anomaly detection[J]. International Journal of Computer Vision, 2021, 129(4): 1038-1059.
[105] ZOU Y, JEONG J, PEMULA L, et al. SPot-the-difference self-supervised pre-training for anomaly detection and segmentation[C]//Proceedings of the European Conference on Computer Vision. Cham: Springer, 2022: 392-408.