About

I am Siva, working as a Senior Research Engineer for Proscia trying to come up with deep learning / machine learning solutions to diagnose cancer in histopathological images. I am particularly interested in the field of applying bayesian principles to deep learning and its applications to medical domain. Papers on generative models (like GANs, VAEs, normalizing flows, autoregressive models), approximate inference, unsupervised representation learning and gazzillion of those self-supervised learning techniques interest me. I have a special liking for papers that connect seemingly disparate techniques / ideas in machine learning / deep learning (like this one which I came across recently, talks about a connection between score matching and noise contrastive estimation). I also have a soft spot for art created using AI like this brilliant (at the same time creepy) piece created just by turning off neurons from a pre-trained GAN.I completed my masters from Boston University under Brian Kulis.

On my leisure time, I sketch, I listen to ambient music, recently I am into travel cinematic videos on Youtube (My favourites creators are, Kriag Adams and Tim Kellner).