Recent improvements in image resolution and acquisition speed have lead to an explosion of published imaging data in materials microscopy. Intensive human efforts, however, are still required to aggregate such published images for use in data-driven research and decision-making pipelines. EXSCLAIM! is a Python toolkit for automatically extracting figure and caption data from open source scientific documents at high volume, as well as for constructing self-annotated materials imaging datasets.
Knowledge of atomistic structure is essential for understanding the properties of materials. While image simulation can facilitate a better understanding of the images in these instances, typical workflows for constructing atomistic models from images are manual and require expert discretion when identifying/interpreting individual features. The ingrained
toolkit provides an automated way to generate an initial guess for the atomistic structure that balances geometrical constraints with a simulated "fit-to-image".
Latent variable and evolutionary modeling techniques to personalize 3D virtual auditory space. Under Construction!