I am currently a computer vision engineer at Unitary where I build computer vision models to ensure safe media content online.
Previously, I was research assistant in verification of deep learning working on distributional robustness to semantic perturbations at University of Oxford. You can read our paper here. >
I also worked as a data scientist at Visulytix for almost 2 years working on developing and building deep learning models for healthcare.
During my MSc in Neurotechnology at Imperial College London, I was part of Dr. Anil Bharath's BICV group, working on 3D unsupervised deep learning models.
My main interests are in AI safety & explainability, unsupervised learning, and the underlying mechanisms of creativity, artificial or not :).
As part of my MSc Thesis, I trained a 3D autoencoder to learn deep representations of the MRI data.
Github RepoGenerated MRI images using a generative adversarial model. Further info here.
Github RepoThis is an ongoing personal project, where I basically have some fun applying latest GANs.
Coming soon...I modelled a 3D fractal phantom to resemble vasculature in wave experiments.
Coming soon...Building computer vision models to detect undesirable content in images and videos.
Working on exploring distributional robustness to generated semantic perturbations using GANs.
Building & implementing state-of-the-art deep learning models for 3D retinal imagery.
Trained a 2D and 3D autoencoder to learn deep representations of the MRI data. Masters Thesis here.
Designed a 3D network of fractal trees in C++ for MagneticResonance Elastography experiments to model arterial branching.
Research Thesis: “Training unsupervised deep learning algorithmson 2D and 3D medical data”.
Research project: “Inferring micro-structural information frommacroscopic elastic parameters determined from shear wavescattering in fractal-like media”.
National Baccalaureate Diploma: 97.3% Overall, including Mathematics and Physics.