I am currently a senior research scientist at NVIDIA, Deep Learning Research.
My vision is transformation of medicine using machine learning. Adoption of AI in medicine however faces challenges such as label scarcity, interpretability, and robustness. The overarching theme of my research is to address these challenges by leveraging tools from statistical signal processing and optimization. I have delivered several contributions to large-scale computational biomedical imaging and sensing using deep generative models; among which I pioneered GANCS for perceptual fast MR imaging.
Prior to that I was a visiting researcher with RISE Lab, EECS Dept., UC Berkeley working with Michael Mahoney.
I received my Ph.D. in EE and Math (minor) from University of Minnesota, Twin Cities (May 2015), under the supervision of Georgios B. Giannakis. My PhD was devoted to large-scale data science for facilitating machine learning. For contributions to large-scale imputation of streaming data I received the Young Author Best Paper Award from IEEE Signal Processing Society 2017.
Dec 2020: 3 papers presented at NeurIPS; “Neural FFTs for universal texture synthesis”; “Risk quantifcation for deep MRI”“ (oral); ”Learning to sense via variational information maximization"
Dec 2020: 4 abstracts submitted to ISMRM 2021
Nov 2020: our paper on “Convex regularization behind neural reconstruction” appeared on arXiv
Nov 2020: delivered a talk about “Sample efficienct learning via self training,” NVIDIA research
Sep 2020: our paper “Uncertainty Quantification in Deep MRI Reconstruction” accepted for IEEE Trans. Medical Imaging (impact factor: 6.7)
Sep 2020: our paper “Neural FFTs for universal texture synthesis” accepted for NeurIPS 2020
Aug 2020: our paper “Wasserstein GANs for MR Imaging: from Paired to Unpaired Training” accepted for IEEE Trans. Medical Imaging (impact factor: 6.7)