Real-IAD D³: A Real-World 2D/Pseudo-3D/3D Dataset for Industrial Anomaly Detection

realiad_1024: This is a version with a resolution of 1024 &times; 1024 (downsampled from realiad_raw), with a manageable size of around 53GB. It is tailored to meet the needs of most researchers.
realiad_raw: This is the raw, high-resolution data, with a size of approximately 507GB.
realiad_jsons.zip: This file consists of data splits in JSON format, providing a structured and organized way to access specific portions of the dataset as needed.

Wenbing Zhu^1*, Lidong Wang^1*, Ziqing Zhou^1*, Chengjie Wang^2,3*, Yurui Pan¹, Ruoyi Zhang⁴, Zhuhao Chen¹, Linjie Cheng¹, Bin-Bin Gao³, Jiangning Zhang³, Zhenye Gan³, Yuxie Wang⁶, Yulong Chen², Shuguang Qian⁴, Mingmin Chi^1†, Bo Peng^5†, Lizhuang Ma^2†,

¹Fudan University ²Shanghai Jiao Tong University ³Youtu Lab, Tencent ⁴Rongcheer Co., Ltd ⁵Shanghai Ocean University ⁶Suzhou University

Coming soon ...