现代信息科技2025，Vol.9Issue(19)：26-30,5.DOI:10.19850/j.cnki.2096-4706.2025.19.006
基于Vision Transformer的混合型晶圆图缺陷模式识别
Defect Pattern Recognition of Mixed Wafer Map Based on Vision Transformer
摘要
Abstract
Wafer testing is an important part of the chip production process.The identification and classification of wafer map defect patterns play a key role in improving the front-end manufacturing process.In the actual production process,various defects may appear at the same time,forming a mixed defect type.The traditional Deep Learning method has a low recognition rate for mixed wafer map defect information.Therefore,this paper proposes a defect recognition method based on Vision Transformer.This method uses the multi-head self-attention mechanism to encode the global features of the wafer map and realizes the efficient identification of mixed wafer defect maps.The experimental results on the mixed defect dataset show that the performance of this method is better than that of the existing Deep Learning model,and the average accuracy is 96.2%.关键词
计算机视觉/晶圆图/缺陷识别/Vision TransformerKey words
computer vision/wafer map/defect recognition/Vision Transformer分类
信息技术与安全科学引用本文复制引用
李攀,娄莉..基于Vision Transformer的混合型晶圆图缺陷模式识别[J].现代信息科技,2025,9(19):26-30,5.
