《计算机应用》唯一官方网站 ›› 2024, Vol. 44 ›› Issue (8): 2595-2603.DOI: 10.11772/j.issn.1001-9081.2023081122
• 多媒体计算与计算机仿真 • 上一篇
收稿日期:
2023-08-22
修回日期:
2023-10-30
接受日期:
2023-11-03
发布日期:
2024-08-22
出版日期:
2024-08-10
通讯作者:
魏伟波
作者简介:
邓凯丽(1992—),女,山东昌乐人,工程师,硕士研究生,主要研究方向:计算机视觉、缺陷检测基金资助:
Kaili DENG, Weibo WEI(), Zhenkuan PAN
Received:
2023-08-22
Revised:
2023-10-30
Accepted:
2023-11-03
Online:
2024-08-22
Published:
2024-08-10
Contact:
Weibo WEI
About author:
DENG Kaili, born in 1992, M. S. candidate, engineer. Her research interests include computer vision, defect detection.Supported by:
摘要:
针对目前只需正常样本即可实现缺陷检测的方法存在漏检或过度检测的问题,构建一种改进掩码自编码器与改进Unet结合的方法实现像素级缺陷检测。首先,采用拟合缺陷模块生成缺陷掩码图像及正常图像对应的缺陷图像;其次,对缺陷图像随机掩码,去除缺陷图像大部分的缺陷信息,激励Transformer结构的自编码器从未掩码的正常区域学习表示并依据上下文修复缺陷图像,为了提高模型对细节的修复能力,设计了新的损失函数;最后,将缺陷图像与修复图像拼接后输入拥有通道方向交叉融合Transformer结构的Unet,实现像素级缺陷检测。实验结果表明,在MVTec AD数据集上,所提方法平均的基于图像的和基于像素的接受者操作特征曲线下的面积值(ROC AUC)分别达到了0.984和0.982,与DRAEM(Discriminatively trained Reconstruction Anomaly Embedding Model)相比分别提高了2.9和3.2个百分点;与CFLOW-AD(Anomaly Detection via Conditional normalizing FLOWs)相比分别提高了3.1和0.8个百分点,证明所提方法具有较高的识别率和检测精度。
中图分类号:
邓凯丽, 魏伟波, 潘振宽. 改进掩码自编码器的工业缺陷检测方法[J]. 计算机应用, 2024, 44(8): 2595-2603.
Kaili DENG, Weibo WEI, Zhenkuan PAN. Industrial defect detection method with improved masked autoencoder[J]. Journal of Computer Applications, 2024, 44(8): 2595-2603.
对象 | 类别 | FAVAE | PaDiM | DFR | RIAD | DRAEM | CFLOW-AD | MemSeg | 本文方法 |
---|---|---|---|---|---|---|---|---|---|
对象 类别 | bottle | (0.991,0.959) | ( | (0.918,0.961) | (0.971,—) | (0.990,0.970) | (1.000,0.981) | (0.995,0.949) | (0.977,0.987) |
metal_nut | (0.925,0.955) | (0.922,0.968) | (0.952,0.896) | (0.920,—) | (0.926, | ( | (0.890,0.865) | (0.989,0.993) | |
cable | (0.798,0.966) | (0.915,0.986) | (0.736, | (0.655,—) | (0.872,0.926) | (0.957,0.988) | ( | (0.945,0.968) | |
capsule | (0.968,0.982) | (0.933,0.979) | (0.981,0.985) | (0.731,—) | ( | (0.999, | (0.937,0.928) | (0.962,0.973) | |
hazelnut | (0.914,0.968) | ( | (0.970,0.951) | (0.890,—) | (0.988, | (0.976,0.982) | (0.921,0.861) | (1.000,0.996) | |
pill | (0.768,0.934) | (0.944,0.961) | (0.952,0.976) | (0.786,—) | ( | (0.938,0.990) | (0.822,0.888) | (0.977, | |
toothbrush | (0.956,0.987) | (0.972,0.987) | ( | (0.972,—) | (1.000,0.981) | (0.917, | (0.972,0.950) | (1.000,0.981) | |
screw | (0.887, | (0.844,0.983) | ( | (0.799,—) | (0.844,0.919) | (0.765,0.971) | (0.914,0.619) | (0.941,0.969) | |
transistor | (0.930,0.959) | (0.978,0.975) | (0.828,0.814) | (0.918,—) | (0.949,0.843) | (0.942,0.904) | (0.887,0.727) | ( | |
zipper | (0.852,0.906) | (0.909, | (0.926,0.975) | (0.975,—) | (0.994,0.945) | (0.988,0.979) | ( | (0.998,0.986) | |
Obj_Aver | (0.899,0.960) | (0.941, | (0.918,0.952) | (0.862,—) | ( | (0.944,0.974) | (0.929,0.865) | (0.976,0.980) | |
纹理 类别 | carpet | (0.904,0.967) | (0.999, | (0.974,0.987) | (0.781,—) | (0.839,0.844) | (0.992,0.992) | (0.902,0.911) | ( |
grid | (0.970, | (0.957,0.965) | (0.887,0.973) | (0.983,—) | ( | (0.886,0.963) | (0.931,0.720 ) | (1.000,0.986) | |
leather | (0.856,0.980) | (1.000,0.989) | ( | (1.000,—) | (1.000, | (1.000,0.996) | (1.000,0.946) | (1.000,0.994) | |
tile | (0.805,0.714) | (0.974,0.939) | (0.938,0.916) | (0.997,—) | (0.985, | ( | (1.000,0.975) | (1.000,0.991) | |
wood | (0.961,0.884) | (0.988,0.941) | (0.986, | (0.965,—) | ( | (0.990,0.941) | (0.990,0.901) | (1.000,0.989) | |
Tex_Aver | (0.899,0.908) | ( | (0.953,0.965) | (0.945,—) | (0.963,0.948) | (0.973, | (0.965,0.891) | (0.999,0.985) | |
平均值 | (0.899,0.943) | ( | (0.930,0.956) | (0.890,—) | ( | (0.953, | (0.941,0.874) | (0.984,0.982) |
表1 各方法异常检测结果对比
Tab. 1 Comparison of abnormaly detection results among different methods
对象 | 类别 | FAVAE | PaDiM | DFR | RIAD | DRAEM | CFLOW-AD | MemSeg | 本文方法 |
---|---|---|---|---|---|---|---|---|---|
对象 类别 | bottle | (0.991,0.959) | ( | (0.918,0.961) | (0.971,—) | (0.990,0.970) | (1.000,0.981) | (0.995,0.949) | (0.977,0.987) |
metal_nut | (0.925,0.955) | (0.922,0.968) | (0.952,0.896) | (0.920,—) | (0.926, | ( | (0.890,0.865) | (0.989,0.993) | |
cable | (0.798,0.966) | (0.915,0.986) | (0.736, | (0.655,—) | (0.872,0.926) | (0.957,0.988) | ( | (0.945,0.968) | |
capsule | (0.968,0.982) | (0.933,0.979) | (0.981,0.985) | (0.731,—) | ( | (0.999, | (0.937,0.928) | (0.962,0.973) | |
hazelnut | (0.914,0.968) | ( | (0.970,0.951) | (0.890,—) | (0.988, | (0.976,0.982) | (0.921,0.861) | (1.000,0.996) | |
pill | (0.768,0.934) | (0.944,0.961) | (0.952,0.976) | (0.786,—) | ( | (0.938,0.990) | (0.822,0.888) | (0.977, | |
toothbrush | (0.956,0.987) | (0.972,0.987) | ( | (0.972,—) | (1.000,0.981) | (0.917, | (0.972,0.950) | (1.000,0.981) | |
screw | (0.887, | (0.844,0.983) | ( | (0.799,—) | (0.844,0.919) | (0.765,0.971) | (0.914,0.619) | (0.941,0.969) | |
transistor | (0.930,0.959) | (0.978,0.975) | (0.828,0.814) | (0.918,—) | (0.949,0.843) | (0.942,0.904) | (0.887,0.727) | ( | |
zipper | (0.852,0.906) | (0.909, | (0.926,0.975) | (0.975,—) | (0.994,0.945) | (0.988,0.979) | ( | (0.998,0.986) | |
Obj_Aver | (0.899,0.960) | (0.941, | (0.918,0.952) | (0.862,—) | ( | (0.944,0.974) | (0.929,0.865) | (0.976,0.980) | |
纹理 类别 | carpet | (0.904,0.967) | (0.999, | (0.974,0.987) | (0.781,—) | (0.839,0.844) | (0.992,0.992) | (0.902,0.911) | ( |
grid | (0.970, | (0.957,0.965) | (0.887,0.973) | (0.983,—) | ( | (0.886,0.963) | (0.931,0.720 ) | (1.000,0.986) | |
leather | (0.856,0.980) | (1.000,0.989) | ( | (1.000,—) | (1.000, | (1.000,0.996) | (1.000,0.946) | (1.000,0.994) | |
tile | (0.805,0.714) | (0.974,0.939) | (0.938,0.916) | (0.997,—) | (0.985, | ( | (1.000,0.975) | (1.000,0.991) | |
wood | (0.961,0.884) | (0.988,0.941) | (0.986, | (0.965,—) | ( | (0.990,0.941) | (0.990,0.901) | (1.000,0.989) | |
Tex_Aver | (0.899,0.908) | ( | (0.953,0.965) | (0.945,—) | (0.963,0.948) | (0.973, | (0.965,0.891) | (0.999,0.985) | |
平均值 | (0.899,0.943) | ( | (0.930,0.956) | (0.890,—) | ( | (0.953, | (0.941,0.874) | (0.984,0.982) |
类别 | A | B | C | NO_CCT | 本文方法 |
---|---|---|---|---|---|
平均值 | (0.966,0.955) | (0.942,0.928) | (0.972,0.962) | (0.980,0.974) | (0.984,0.982) |
bottle | (0.950,0.953) | (0.961,0.979) | (0.960,0.984) | (0.984,0.985) | (0.977,0.987) |
metal_nut | (0.980,0.965) | (0.999,0.992) | (0.992,0.982) | (0.988,0.984) | (0.989,0.993) |
cable | (0.952,0.955) | (0.932,0.934) | (0.933,0.962) | (0.954,0.968) | (0.945,0.968) |
capsule | (0.875,0.892) | (0.946,0.917) | (0.926,0.952) | (0.957,0.957) | (0.962,0.973) |
hazelnut | (0.992,0.962) | (1.000,0.971) | (1.000,0.987) | (1.000,0.989) | (1.000,0.996) |
pill | (0.952,0.954) | (0.827,0.902) | (0.945,0.967) | (0.933,0.978) | (0.977,0.981) |
toothbrush | (1.000,0.957) | (1.000,0.974) | (1.000,0.943) | (1.000,0.953) | (1.000,0.981) |
screw | (0.934,0.921) | (0.900,0.901) | (0.941,0.861) | (0.938,0.965) | (0.941,0.969) |
transistor | (0.961,0.937) | (0.962,0.841) | (0.965,0.934) | (0.975,0.935) | (0.969,0.966) |
zipper | (0.992,0.973) | (1.000,0.972) | (0.999,0.965) | (0.996,0.963) | (0.998,0.986) |
carpet | (0.945,0.961) | (0.846,0.939) | (0.935,0.973) | (0.982,0.967) | (0.996,0.964) |
grid | (0.972,0.971) | (0.830,0.770) | (0.992,0.984) | (1.000,0.989) | (1.000,0.986) |
leather | (1.000,0.986) | (1.000,0.987) | (1.000,0.982) | (1.000,0.993) | (1.000,0.994) |
tile | (1.000,0.981) | (0.997,0.981) | (1.000,0.988) | (1.000,0.992) | (1.000,0.991) |
wood | (0.983,0.954) | (0.932,0.857) | (0.991,0.967) | (1.000,0.991) | (1.000,0.989) |
表2 不同实验条件的检测结果
Tab. 2 Detection results under different experimental conditions
类别 | A | B | C | NO_CCT | 本文方法 |
---|---|---|---|---|---|
平均值 | (0.966,0.955) | (0.942,0.928) | (0.972,0.962) | (0.980,0.974) | (0.984,0.982) |
bottle | (0.950,0.953) | (0.961,0.979) | (0.960,0.984) | (0.984,0.985) | (0.977,0.987) |
metal_nut | (0.980,0.965) | (0.999,0.992) | (0.992,0.982) | (0.988,0.984) | (0.989,0.993) |
cable | (0.952,0.955) | (0.932,0.934) | (0.933,0.962) | (0.954,0.968) | (0.945,0.968) |
capsule | (0.875,0.892) | (0.946,0.917) | (0.926,0.952) | (0.957,0.957) | (0.962,0.973) |
hazelnut | (0.992,0.962) | (1.000,0.971) | (1.000,0.987) | (1.000,0.989) | (1.000,0.996) |
pill | (0.952,0.954) | (0.827,0.902) | (0.945,0.967) | (0.933,0.978) | (0.977,0.981) |
toothbrush | (1.000,0.957) | (1.000,0.974) | (1.000,0.943) | (1.000,0.953) | (1.000,0.981) |
screw | (0.934,0.921) | (0.900,0.901) | (0.941,0.861) | (0.938,0.965) | (0.941,0.969) |
transistor | (0.961,0.937) | (0.962,0.841) | (0.965,0.934) | (0.975,0.935) | (0.969,0.966) |
zipper | (0.992,0.973) | (1.000,0.972) | (0.999,0.965) | (0.996,0.963) | (0.998,0.986) |
carpet | (0.945,0.961) | (0.846,0.939) | (0.935,0.973) | (0.982,0.967) | (0.996,0.964) |
grid | (0.972,0.971) | (0.830,0.770) | (0.992,0.984) | (1.000,0.989) | (1.000,0.986) |
leather | (1.000,0.986) | (1.000,0.987) | (1.000,0.982) | (1.000,0.993) | (1.000,0.994) |
tile | (1.000,0.981) | (0.997,0.981) | (1.000,0.988) | (1.000,0.992) | (1.000,0.991) |
wood | (0.983,0.954) | (0.932,0.857) | (0.991,0.967) | (1.000,0.991) | (1.000,0.989) |
类别 | maskRatio值 | ||
---|---|---|---|
0.25 | 0.50 | 0.75 | |
平均值 | (0.938,0.939) | (0.968,0.959) | (0.973,0.968) |
carpet | (0.880,0.946) | (0.974,0.953) | (0.982,0.967) |
wood | (0.991,0.964) | (0.998,0.979) | (1.000,0.991) |
transistor | (0.939,0.859) | (0.972,0.889) | (0.975,0.935) |
cable | (0.938,0.961) | (0.947,0.972) | (0.954,0.968) |
metal_nut | (0.978,0.963) | (0.985,0.991) | (0.988,0.984) |
screw | (0.902,0.941) | (0.932,0.968) | (0.938,0.965) |
表3 不同掩码比例的异常检测结果
Tab. 3 Defect detection results with different mask ratios
类别 | maskRatio值 | ||
---|---|---|---|
0.25 | 0.50 | 0.75 | |
平均值 | (0.938,0.939) | (0.968,0.959) | (0.973,0.968) |
carpet | (0.880,0.946) | (0.974,0.953) | (0.982,0.967) |
wood | (0.991,0.964) | (0.998,0.979) | (1.000,0.991) |
transistor | (0.939,0.859) | (0.972,0.889) | (0.975,0.935) |
cable | (0.938,0.961) | (0.947,0.972) | (0.954,0.968) |
metal_nut | (0.978,0.963) | (0.985,0.991) | (0.988,0.984) |
screw | (0.902,0.941) | (0.932,0.968) | (0.938,0.965) |
1 | TAO X, ZHANG D, WANG Z, et al. Detection of power line insulator defects using aerial images analyzed with convolutional neural networks [J]. IEEE Transactions on Systems, Man, and Cybernetics: Systems, 2020, 50(4): 1486-1498. |
2 | HE Y, SONG K, MENG Q, et al. An end-to-end steel surface defect detection approach via fusing multiple hierarchical features [J]. IEEE Transactions on Instrumentation and Measurement, 2020, 69(4): 1493-1504. |
3 | LI J, SU Z, GENG J, et al. Real-time detection of steel strip surface defects based on improved YOLO detection network [J]. IFAC-PapersOnLine, 2018, 51(21): 76-81. |
4 | 刘艳菊, 王秋霁, 赵开峰, 等. 基于卷积神经网络的热轧钢条表面实时缺陷检测 [J]. 仪器仪表学报, 2021, 42(12): 211-219. |
LIU Y J, WANG Q J, ZHAO K F, et al. Real-time defect detection of hot rolling steel bar based on convolution neural network [J]. Chinese Journal of Scientific Instrument, 2021, 42(12): 211-219. | |
5 | 陈仁祥, 詹赞, 胡小林, 等. 基于多注意力Faster RCNN的噪声干扰下印刷电路板缺陷检测[J]. 仪器仪表学报, 2021, 42(12): 167-174. |
CHEN R X, ZHAN Z, HU X L, et al. Printed circuit board defect detection based on the multi-attentive faster RCNN under noise interference [J]. Chinese Journal of Scientific Instrument, 2021, 42(12): 167-174. | |
6 | TAO X, GONG X, 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. |
7 | BAUR C, WIESTLER B, ALBARQOUNI S, et al. Deep autoencoding models for unsupervised anomaly segmentation in brain MR image [C]// Proceedings of the 4th International Workshop on Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries. Cham: Springer, 2019: 161-169. |
8 | SCHLEGL T, SEEBÖCK P, WALDSTEIN S M, et al. Unsupervised anomaly detection with Generative Adversarial Networks to guide marker discovery [C]// Proceedings of the 25th International Conference on Information Processing in Medical Imaging. Cham: Springer, 2017: 146-157. |
9 | 唐善成, 陈明, 王瀚博, 等. 采用变分自编码器的无监督压敏电阻表面缺陷检测 [J]. 计算机集成制造系统, 2022, 28(5): 1337-1351. |
TANG S C, CHEN M, WANG H B, et al. Unsupervised varistor surface defect detection based on variational autoencoder [J]. Computer Integrated Manufacturing Systems, 2022, 28(5): 1337-1351. | |
10 | 余文勇, 张阳, 姚海明, 等. 基于轻量化重构网络的表面缺陷视觉检测 [J]. 自动化学报, 2022, 48(9): 2175-2186. |
YU W Y, ZHANG Y, YAO H M, et al. Visual inspection of surface defects based on lightweight reconstruction network [J]. Acta Automatica Sinica, 2022, 48(9): 2175-2186. | |
11 | SCHLEGL T, SEEBÖCK P, WALDSTEIN S M, et al. f-AnoGAN: fast unsupervised anomaly detection with generative adversarial networks [J]. Medical Image Analysis, 2019, 54: 30-44. |
12 | AKCAY S, ATAPOUR-ABARGHOUEI A, BRECKON T P. GANomaly: semi-supervised anomaly detection via adversarial training [C]// Proceedings of the14th Asian Conference on Computer Vision. Cham: Springer, 2019: 622-637. |
13 | AKÇAY S, ATAPOUR-ABARGHOUEI A, BRECKON T P. Skip-GANomaly: skip connected and adversarially trained encoder-decoder anomaly detection [C]// Proceedings of the 2019 International Joint Conference on Neural Networks. Piscataway: IEEE, 2019: 1-8. |
14 | TANG T-W, HSU H, HUANG W-R, et al. Industrial anomaly detection with skip autoencoder and deep feature extractor [J]. Sensors, 2022, 22(23): 9327-9338. |
15 | SHI Y, YANG J, QI Z. Unsupervised anomaly segmentation via deep feature reconstruction [J]. Neurocomputing, 2021, 424: 9-22. |
16 | DEHAENE D, ELINE P. Anomaly localization by modeling perceptual features [EB/OL]. (2020-08-12) [2023-06-20]. . |
17 | MISHRA P, VERK R, FORNASIER D, et al. VT-ADL: a vision transformer network for image anomaly detection and localization [C]// Proceedings of the 2021 IEEE 30th International Symposium on Industrial Electronics. Piscataway: IEEE, 2021: 1-6. |
18 | PIRNAY J, CHAI K. Inpainting transformer for anomaly detection [C]// Proceedings of the 21st International Conference on Image Analysis and Processing. Cham: Springer, 2022: 394-406. |
19 | WAN Q, GAO L, LI X, et al. Unsupervised image anomaly detection and segmentation based on pretrained feature mapping [J]. IEEE Transactions on Industrial Informatics, 2023, 19(3): 2330-2339. |
20 | HASELMANN M, GRUBER D P, TABATABAI P. Anomaly detection using deep learning based image completion [C]// Proceedings of the 2018 17th IEEE International Conference on Machine Learning and Applications. Piscataway: IEEE, 2018: 1237-1242. |
21 | 张宏伟,谭全露,陆帅,等.U型去噪卷积自编码器色织衬衫裁片缺陷检测[J]. 西安电子科技大学学报, 2021, 48(3):123-130. |
ZHANG H W, TAN Q L, LU S, et al. Yarn-dyed shirt piece defect detection based on an unsupervised reconstruction model of the U-shaped denoising convolutional auto-encoder [J]. Journal of Xidian University, 2021, 48(3): 123-130. | |
22 | ZAVRTANIK V, KRISTAN M, SKOČAJ D. Reconstruction by inpainting for visual anomaly detection [J]. Pattern Recognition, 2021, 112: 107706. |
23 | JIANG J, ZHU J, BILAL M,et al. Masked Swin Transformer Unet for industrial anomaly detection [J]. IEEE Transactions on Industrial Informatics, 2023, 19(2): 2200-2209. |
24 | DE OLIVEIRA D C, NASSU B T, WEHRMEISTER M A, et al. Image-based detection of modifications in assembled PCBs with deep convolutional autoencoders [J]. Sensors, 2023, 23(3): 1353. |
25 | TAO X, ZHANG D, MA W, et al. Unsupervised anomaly detection for surface defects with dual-Siamese network [J]. IEEE Transactions on Industrial Informatics, 2022, 18(11): 7707-7717. |
26 | YANG H, ZHOU Q, SONG K, et al. An anomaly feature-editing-based adversarial network for texture defect visual inspection [J]. IEEE Transactions on Industrial Informatics, 2021, 17(3): 2220-2230. |
27 | 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. | |
28 | ZHANG Z, DENG X. Anomaly detection using improved deep SVDD model with data structure preservation [J]. Pattern Recognition Letters, 2021, 148: 1-6. |
29 | YI J, YOON S. Patch SVDD: patch-level SVDD for anomaly detection and segmentation [EB/OL]. (2020-07-13) [2023-06-20]. . |
30 | LI C-L, SOHN K, YOON J, et al. CutPaste: Self-supervised learning for anomaly detection and localization [C]// Proceedings of the 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2021: 9659-9669. |
31 | C-C TSAI, WU T-H, LAI S-H, et al. Multi-scale patch-based representation learning for image anomaly detection and segmentation [C]// Proceedings of the 2022 IEEE/CVF Winter Conference on Applications of Computer Vision. Piscataway: IEEE, 2022: 3065-3073. |
32 | COHEN N, HOSHEN Y. Sub-image anomaly detection with deep pyramid correspondences [EB/OL]. (2021-02-03) [2023-06-20]. . |
33 | DEFARD T, SETKOV A, LOESCH A, et al. PaDiM: a patch distribution modeling framework for anomaly detection and localization [EB/OL]. (2020-11-17) [2023-06-20]. . |
34 | 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. |
35 | GUDOVSKIY D, ISHIZAKA S, KOZUKA K, et al. CFLOW-AD: real-time unsupervised anomaly detection with localization via conditional normalizing flows [C]// Proceedings of the 22nd IEEE/CVF Winter Conference on Applications of Computer Vision. Piscataway: IEEE, 2022: 1819-1828. |
36 | LEE S, LEE S, SONG B C. CFA: coupled-hypersphere-based feature adaptation for target-oriented anomaly localization [J]. IEEE Access, 2022, 10: 78446-78454. |
37 | YANG M, WU P, FENG H. MemSeg: a semi-supervised method for image surface defect detection using differences and commonalities [J]. Engineering Applications of Artificial Intelligence, 2023, 119: 105835. |
38 | ZAVRTANIK V, KRISTAN M, SKOČAJ D, et al. DRAEM: a discriminatively trained reconstruction embedding for surface anomaly detection [C]// Proceedings of the 2021 IEEE/CVF International Conference on Computer Vision. Piscataway: IEEE, 2021: 8310-8319. |
39 | PERLIN K. An image synthesizer [J]. ACM SIGGRAPH Computer Graphics, 1985, 19(3): 287-296. |
40 | CUBUK E D, ZOPF B, SHLENS J, et al. Randaugment: practical automated data augmentation with a reduced search space [C]// Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE,2020: 3008-3017. |
41 | HE K, CHEN X, XIE S,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. |
42 | WANG H, CAO P, WANG J, et al. UCTransNet: Rethinking the skip connections in U-Net from a channel-wise perspective with transformer [J]. Proceedings of the AAAI Conference on Artificial Intelligence, 2022, 36(3): 2441-2449. |
43 | LIN T Y, GOYAL P, GIRSHICK R, et al. Focal loss for dense object detection [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2020, 42(2): 318-327. |
44 | 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. |
[1] | 丁宇伟, 石洪波, 李杰, 梁敏. 基于局部和全局特征解耦的图像去噪网络[J]. 《计算机应用》唯一官方网站, 2024, 44(8): 2571-2579. |
[2] | 李大海, 王忠华, 王振东. 结合空间域和频域信息的双分支低光照图像增强网络[J]. 《计算机应用》唯一官方网站, 2024, 44(7): 2175-2182. |
[3] | 黎施彬, 龚俊, 汤圣君. 基于Graph Transformer的半监督异配图表示学习模型[J]. 《计算机应用》唯一官方网站, 2024, 44(6): 1816-1823. |
[4] | 黄梦源, 常侃, 凌铭阳, 韦新杰, 覃团发. 基于层间引导的低光照图像渐进增强算法[J]. 《计算机应用》唯一官方网站, 2024, 44(6): 1911-1919. |
[5] | 吕锡婷, 赵敬华, 荣海迎, 赵嘉乐. 基于Transformer和关系图卷积网络的信息传播预测模型[J]. 《计算机应用》唯一官方网站, 2024, 44(6): 1760-1766. |
[6] | 刘子涵, 周登文, 刘玉铠. 基于全局依赖Transformer的图像超分辨率网络[J]. 《计算机应用》唯一官方网站, 2024, 44(5): 1588-1596. |
[7] | 席治远, 唐超, 童安炀, 王文剑. 基于双路时空网络的驾驶员行为识别[J]. 《计算机应用》唯一官方网站, 2024, 44(5): 1511-1519. |
[8] | 孙子文, 钱立志, 杨传栋, 高一博, 陆庆阳, 袁广林. 基于Transformer的视觉目标跟踪方法综述[J]. 《计算机应用》唯一官方网站, 2024, 44(5): 1644-1654. |
[9] | 王铂越, 李英祥, 钟剑丹. 基于改进Res-UNet的昼夜地基云图分割网络[J]. 《计算机应用》唯一官方网站, 2024, 44(4): 1310-1316. |
[10] | 黄荣, 宋俊杰, 周树波, 刘浩. 基于自监督视觉Transformer的图像美学质量评价方法[J]. 《计算机应用》唯一官方网站, 2024, 44(4): 1269-1276. |
[11] | 吴宁, 罗杨洋, 许华杰. 基于多尺度特征融合的遥感图像语义分割方法[J]. 《计算机应用》唯一官方网站, 2024, 44(3): 737-744. |
[12] | 黄巧玲, 郑伯川, 丁梓成, 吴泽东. 融合监督注意力模块和跨阶段特征融合的图像修复改进网络[J]. 《计算机应用》唯一官方网站, 2024, 44(2): 572-579. |
[13] | 崔晨辉, 蔺素珍, 李大威, 禄晓飞, 武杰. 基于孪生网络和Transformer的红外弱小目标跟踪方法[J]. 《计算机应用》唯一官方网站, 2024, 44(2): 563-571. |
[14] | 陈丽安, 过弋. 融合个体偏差信息的文本情感分析模型[J]. 《计算机应用》唯一官方网站, 2024, 44(1): 145-151. |
[15] | 陈豪, 夏振平, 程成, 林李兴, 张博文. 基于Transformer-CNN的轻量级图像超分辨率重建网络[J]. 《计算机应用》唯一官方网站, 2024, 44(1): 292-299. |
阅读次数 | ||||||
全文 |
|
|||||
摘要 |
|
|||||