Machine Learning for Sub-pixel Super-resolution in Direct Electron Detectors

Direct electron detectors (DED) are now able to image biological specimens with higher resolution than x-rays—and are enabling new materials imaging techniques such as 4D-STEM. Limitations imposed by semiconductor manufacturing processes, where consumer electronics dominate device chip surface area, reduce the available space for DED pixels. To push beyond the physical pixel, sub-pixel super-resolution is necessary. We explore the prospects for sub-pixel super-resolution through electron counting as a function of diode depth, pixel pitch, and beam energy. For most energy ranges of interest to electron microscopists, energy is deposited as a string of charge across multiple pixels. We use machine learning to identify the start of the string, determining the true entry point of an electron with greater success than existing electron counting statistics.