Extraction and Reconstruction of Data Points from Computer aided Design Using Deep Learning

Variational autoencoders (VAEs) highly important in the fields of machine learning and artificial intelligence because of their versatility and effectiveness in various applications. The current study aims to investigate the application of deep learning models to reduce dimensionality and extract valuable properties from Computer-Aided Design (CAD) models. With an emphasis on the shaft model, this research study described a novel method for data compression and reconstruction of the CAD model using VAEs. The Python programming language is used in the study, together with its numerous libraries and frameworks for deep learning and data processing. The suggested method aims to reduce storage requirements, improve data accessibility, and preserve the integrity of the original CAD models, making an important contributions to the disciplines of CAD, data compression, and artificial intelligence. The findings of the research suggested that a variational autoencoder with a rectified linear unit (ReLU) provides a better correlation of approximately 95% between extracted and constructed data.