Baskar Ganapathysubramanian

TL;DR: It is shown that a deep learning architecture developed for rare object identification in clutter-filled images can identify and count the SCN eggs and illustrates the remarkable promise of applying deep learning approaches to phenotyping for pest assessment and management.

TL;DR: A novel approach that employs IGA methodologies while still rigorously abiding by the paradigms of advanced design parameterization, analysis model validity, and interactivity is proposed, demonstrating the framework’s effectiveness on both an internally pressurized tube and a wind turbine blade.

TL;DR: It is proposed that deep learning methods can completely outpace current approaches for scientific inverse problems while delivering comparable designs, and it is shown how intelligent sampling of the design space inputs can makeDeep learning methods more competitive in accuracy, while illustrating their generalization capability to out-of-sample predictions.

TL;DR: A “transparent soil” formed by the spherification of hydrogels of biopolymers is described with high transparency, good mechanical stability, tunable pore sizes, low cost, and easy scalability, and allows for the imaging of unconstrained root systems in vivo by both photography and microscopy.

TL;DR: This work forms a reinforcement learning (RL)-based design framework that mitigates disadvantages of both optimization and supervised learning approaches, and shows that a single generic RL agent is capable of exploring the solution space to achieve multiple design objectives.