Chen Y.-SHUI-RU JIANG2023-06-092023-06-092022https://www.scopus.com/inward/record.uri?eid=2-s2.0-85130776524&doi=10.23919%2fDATE54114.2022.9774622&partnerID=40&md5=1a88ee5d48bd1bfb8f970cfd6572d8c3https://scholars.lib.ntu.edu.tw/handle/123456789/632486Exploring hotspot patterns and correcting them as early as possible is crucial to guarantee yield and manufacturability. Existing hotspot detection and pattern classification methods consider only the geometry on one single layer or one main layer with adjacent layers. In this paper, we investigate the linkage between many-layer hotspot patterns and potentially induced defect types. We first cast the many-layer critical hotspot pattern extraction task as a visual question answering (VQA) problem: Considering a many-layer layout pattern an image and a defect type a question, we devise a layer-attentioned VQA model to answer whether the pattern is critical to the queried defect type. Furthermore, our layer attention mechanism attempts to identify the relevance of each layer for different defect types. Experimental results demonstrate that the proposed model has superior question-answering ability for modern layouts with more than thirty layout layers. © 2022 EDAA.Defects; Adjacent layers; Classification methods; Defect type; Hotspot detections; Hotspots; Induced defects; Manufacturability; Patterns classification; Question Answering; Single layer; Pattern recognitionconference paper10.23919/DATE54114.2022.97746222-s2.0-85130776524