收藏切换
DiffAD: Difference Convolution Attention-Based Multi-Class Unsupervised Anomaly Detection
收藏切换
Jian Yao1, 2, Shuai Ma1, 2, Pengjiang Qian1, 2, *, Chuang Wang1, 2, Guisong Yang3, Yuan He4, 5
Journal of Computer Research and Development | 2026, 63(8) : 2111 - 2126
Less
收藏切换
Journal of Computer Research and Development | 2026, 63(8): 2111-2126
• Artificial Intelligence •
DiffAD: Difference Convolution Attention-Based Multi-Class Unsupervised Anomaly Detection
Full
Jian Yao1, 2, Shuai Ma1, 2, Pengjiang Qian1, 2, *, Chuang Wang1, 2, Guisong Yang3, Yuan He4, 5
Affiliations
  • 1School of Artificial Intelligence and Computer Science (School of Software), Jiangnan University, Wuxi, Jiangsu 214122
  • 2Engineering Research Center of Intelligent Technology for Healthcare (Jiangnan University), Ministry of Education, Wuxi, Jiangsu 214122
  • 3School of Optical-Electrical and Computer Engineering, University of Shanghai for Science and Technology, Shanghai 200093
  • 4School of Software, Tsinghua University, Beijing 100084
  • 5Beijing National Research Center for Information Science and Technology (Tsinghua University), Beijing 100084
  • born in 1984. Master, senior engineer. His main research interests include lightweight edge learning, explainable artificial intelligence, and smart healthcare

    born in 2001. Bachelor. His main research interests include deep learning and product defect detection

    born in 1979. PhD, professor, PhD supervisor. Senior member of CCF. His main research interests include machine learning,deep learning, intelligent healthcare, and smart city

    born in 1998. Master. His main research interests include machine learning, deep learning, and explainable artificial intelligence

    born in 1982. PhD, associate professor, master supervisor. His main research interests include internet of things and ubiquitous computing, mobile crowd-sensing, and edge computing

    born in 1980. PhD, professor, PhD supervisor. Distinguished member of CCF. His main research interests include internet of things, wireless networks, and mobile and ubiquitous computing

About Author:

born in 1984. Master, senior engineer. His main research interests include lightweight edge learning, explainable artificial intelligence, and smart healthcare

born in 2001. Bachelor. His main research interests include deep learning and product defect detection

born in 1979. PhD, professor, PhD supervisor. Senior member of CCF. His main research interests include machine learning,deep learning, intelligent healthcare, and smart city

born in 1998. Master. His main research interests include machine learning, deep learning, and explainable artificial intelligence

born in 1982. PhD, associate professor, master supervisor. His main research interests include internet of things and ubiquitous computing, mobile crowd-sensing, and edge computing

born in 1980. PhD, professor, PhD supervisor. Distinguished member of CCF. His main research interests include internet of things, wireless networks, and mobile and ubiquitous computing

Published: 2026-08-01 doi: 10.7544/issn1000-1239.202550706
Outline
收藏切换

A novel unsupervised multi-class anomaly detection model, termed DiffAD, is proposed to tackle the critical challenges of annotation scarcity and insufficient detection accuracy plaguing visual anomaly detection tasks in complex industrial scenarios. The model employs a progressive feature reconstruction strategy, with its core innovation centered on a dedicated feature reconstruction component named DADE. Breaking new ground, DADE decomposes the feature reconstruction process into three sequential phases, namely chaos, pre-refinement and strong refinement, which synergistically enhance the quality of reconstructed features and ensure the stability of the entire reconstruction workflow. A key standout of DADE lies in its seamless integration of difference convolution attention and a detail enhancement mechanism. It combines difference convolution with multi-head self-attention and incorporates residual dense connections, thereby enabling the component to markedly strengthen the model’s ability to capture subtle image variations and high-frequency details that are critical for anomaly identification and substantially elevate the precision of pixel-level anomaly localization. Extensive empirical evaluations are conducted on four representative benchmark datasets, including MVTec-AD, VisA, MVTec-3D and Uni-Medical. The experimental results consistently demonstrate that DiffAD achieves significant performance superiority over state-of-the-art mainstream models across both image-level and pixel-level anomaly detection metrics, fully underscoring its remarkable practical application value and great potential for widespread deployment in the domain of unsupervised visual anomaly detection.

unsupervised  /  anomaly detection  /  computer vision  /  attention mechanism  /  feature reconstruction
姚健, 马帅, 钱鹏江, 王闯, 杨桂松, 何源. DiffAD:基于差分卷积注意力的多类无监督异常检测. 计算机研究与发展, 2026 , 63 (8) : 2111 -2126 . DOI: 10.7544/issn1000-1239.202550706
Jian Yao, Shuai Ma, Pengjiang Qian, Chuang Wang, Guisong Yang, Yuan He. DiffAD: Difference Convolution Attention-Based Multi-Class Unsupervised Anomaly Detection[J]. Journal of Computer Research and Development, 2026 , 63 (8) : 2111 -2126 . DOI: 10.7544/issn1000-1239.202550706
Year 2026 volume 63 Issue 8
85
37
Cite this Article
BibTeX
Article Info
doi: 10.7544/issn1000-1239.202550706
  • Receive Date:2025-09-21
  • Online Date:2026-11-17
  • Published:2026-08-01
Article Data
Affiliations
History
  • Received:2025-09-21
  • Revised:2026-03-30
  • Accepted:2026-04-07
Affiliations
    1School of Artificial Intelligence and Computer Science (School of Software), Jiangnan University, Wuxi, Jiangsu 214122
    2Engineering Research Center of Intelligent Technology for Healthcare (Jiangnan University), Ministry of Education, Wuxi, Jiangsu 214122
    3School of Optical-Electrical and Computer Engineering, University of Shanghai for Science and Technology, Shanghai 200093
    4School of Software, Tsinghua University, Beijing 100084
    5Beijing National Research Center for Information Science and Technology (Tsinghua University), Beijing 100084
References
Share
https://castjournals.cast.org.cn/joweb/jsjyjyfz/EN/10.7544/issn1000-1239.202550706
Share to
QR

Scan QR to access full text

Cite this article
BibTeX
Citations
表12种不同金属材料的力学参数
科
Family
属数
Number of
genus
种数
Number of
species
占总种数比例
Percentage of
total species (%)
属
Genus
种数
Number of
species
占总种数比例
Percentage of total
species (%)
鹅膏菌科Amanitaceae 2 11 5.26 鹅膏菌属 Amanita 10 4.78
小菇科 Mycenaceae 2 12 5.74 丝盖伞属 Inocybe 5 2.39
多孔菌科 Polyporaceae 8 14 6.70 蜡蘑属 Laccaria 5 2.39
红菇科 Russulaceae 3 23 11.00 小皮伞属 Marasmius 6 2.87
小菇属 Mycena 11 5.26
光柄菇属 Pluteus 5 2.39
红菇属 Russula 17 8.13
栓菌属 Trametes 5 2.39
关闭全屏
  • BibTeX
  • EndNote
  • RefWorks
  • TxT