About

Sonaal Kant is a highly skilled and accomplished Applied Scientist with over 6 years of relevant experience in the field of Computer Science. He is currently pursuing his Master of Science in Computer Science at the University of Maryland while working as an Applied Scientist Intern at Amazon Fashion. In his current role, Sonaal has been working on creating a 3D body model using self-supervision with differentiable rendering from a monocular RGB image. This 3D model is then used to calculate the body measurements of the user from a single RGB image, which are further used to provide a custom fit T-shirt for the user. Sonaal's work is focused on estimating body shape from an image, which is still an open and less explored research problem due to the high variance in Body Mass Index of the population and less available ground truth data. He is also working with self-supervision to overcome these challenges. Prior to his current role, Sonaal worked as a Research Assistant at the University of Maryland, where he worked on the low-shot learning problem for object classification in images by generating new samples for the low-shot categories using Generative Adversarial methods in feature space. He also worked on developing a pipeline for subject-object interaction in images alongside generating a baseline using Spatially Conditioned Graphs and developing capabilities for novelty detection. Sonaal has also worked as a Senior Data Scientist at Paralleldots, where he was involved in the research and pipeline of SmartGaze from scratch. He achieved more than 95% accuracy on videos using many Deep learning models in combination. He also worked on research and end-to-end pipeline for ShelfWatch, where a general product detector and meta classifier were used for real-time and few-shot SKU detection and classification, respectively. Sonaal holds a Master of Science in Computer Science from the University of Maryland and a Bachelor of Technology in Information Technology from Maharaja Agrasen Institute of Technology. He has expertise in Research Science, Data Science, and Deep Learning. Sonaal is a highly motivated and results-driven individual who is passionate about exploring new technologies and finding innovative solutions to complex problems.