Microchips are made by building up layers of interconnected patterns on a silicon wafer. The manufacturing process involves hundreds of steps, done with nanometer precision. At the most cutting-edge, it requires EUV lithography.
New algorithms for reconstruction of wafers from (denoised) SEM images
To further develop advanced lithography, a combination of physical characteristics and data-derived algorithms establishes greater densities of chips. This project develops new efficient algorithms for sub-surface reconstruction of wafers from (denoised) SEM images that include physical models of the sub-surface reconstruction of wafers, and advanced deep generative models for stochastic mask prediction to optimize for worst-case performance (robust statistics) to achieve highly robust, yet performant, masks.