arrow
返回

Context-aware defect sample generation using conditional diffusion model for surface defect inspection

delete2026-04-07
delete0
PRE
AI
C
Changwoo Nam
S
Sang Jun Lee *
DOI:10.1016/j.measurement.2026.121432delete
delete原文链接
delete原文求助
delete分享
delete收藏
摘要

摘要

En 中文
• 一种情境感知扩散方法生成逼真的工业表面缺陷。 • 所提出的SCMP与复杂背景保持结构一致性。 • CMDP有效消除不自然的边界伪影。 • 在钢和织物缺陷数据集上验证了自动化视觉缺陷检测。 • 克服数据稀缺以提高视觉测量系统的可靠性。
Keyword:
context-aware diffusion
structural consistency
boundary artifacts
surface defect inspection
data scarcity

期刊

Measurement 封面图
Measurement
IF:
5.6
论文数:
2.0W
被引数:
5.4W

机构

暂无机构信息
引用论文

引用论文

Simulation Model of a Steelmaking-Continuous Casting Process Based on Dynamic-Operation Rules
err2024-09-03
err0
errOAAI
errShao, Xin; Liu, Qing; Chen, Hongzhi; Zhang, Jiangshan; Gao, Shan; Li, Shaoshuai
err分享
err收藏
Generative data augmentation by conditional inpainting for multi-class object detection in infrared images
err2024-09-01
err1
errOAAI
errWang, Peng; Ma, Zhe; Dong, Bo; Liu, Xiuhua; Ding, Jishiyu; Sun, Kewu; Chen, Ying
err分享
err收藏
学者 查看更多内容