Convolutional neural network model based on terahertz imaging for integrated circuit defect detections

摘要: Detection of integrated circuit (IC) defects is vital in IC manufacturing. Traditional defect detection methods have relied on scanning electron microscopy and X-ray imaging techniques that are time consuming destructive. Hence, this paper we considered terahertz as a label-free nondestructive alternative. This study aimed to use convolutional neural network model (CNN) improve the performance system. First, constructed dataset analyzed it. Subsequently, new CNN structure was proposed based VGG16 network. Finally, it optimized its dropout rate. The method above can accuracy THz imaging. Most significantly, work will promote application practice provide foundation further research relevant fields.