Responsive image

About Me


Hi there! I'm Sanket Lokegaonkar, currently a software engineer at Bloomberg. I graduated from Virginia Tech with Masters in Computer Science. As a masters student, I focused on computer vision research in Prof. Jia-Bin's Vision and Learning Lab . I was jointly advised by Prof. Jia-Bin Huang and Prof. Naren Ramakrishnan . My thesis focused on developing a new continual learning method to alleviate catastrophic forgetting in incremental pixel-wise/dense labeling problems.

Timeline


Fall 2018 : Joined Bloomberg as a Software Engineer!!

Spring 2018 : Graduated!!! Woohoo!! Thesis : Continual Learning for Deep Dense Prediction

Fall 2017 : Research Assistant under Prof. Jia-Bin. Developed the photogrammetry and iterative 3D reconstruction workflow for Source Form, a tool to generate/print 3D models from subjective textual user input.

Fall 2017 : Proposed and developed constrained joint matrix factorization approach for game and user recommendation engine on Steam gaming dataset . Extended the algorithm for MapReduce Framework.

Summer 2017: Research Assistant under Prof. Naren at Discovery Analytics Center and Virginia Tech Transportation Institute. Analyzed and developed methods to detect driver fatigue/drowsiness and identify distracted drivers, using dash-cam video and car sensors.

Spring 2017 : Researched on possible improvements for siamese neural nets based visual tracking in videos with Prof. Jia-Bin Huang.

Spring 2017 : Developed the first prototype for American football quarterback virtual reality/training environment in a camera-tracked space (The Cube), with Prof. Doug Bowman. Checkout the awesome demo here.

Spring 2017 : Teaching assistant for CS 3714: Mobile Software Development at VT.

Fall 2016 : Researched and developed approches to learn deep agents on half-field offense task in a multi-agent soccer environment. The approach builds on the model by He, et al. "Opponent Modeling in Deep Reinforcement Learning". [Keras codebase] [For more info, checkout the thesis by Matthew Hausknecht]

Fall 2016 : Collaborated with Yue Cheng , Pratik Anand , Richa Sinha and Prof. Ali.R.Butt , to develop optimizations in Redis targeted for reduction in memory usage for Redis, a key-value store database. [report ] [code ]

Fall 2016 : Joined Virginia Tech for M.S in Computer Science.

Fall 2015 - Spring 2016 : Research intern at Indian Institute of Technology, Bombay, in collaboration with Ganesh Ramakrishnan. Worked on developing a domain-agnostic framework for building and evolving a taxonomy. Core contributor to development of web application and mobile application (Lokavidya) focusing on content aggregation, collaborative content creation and dissemination of informational multimedia content in rural India.[ Lokacart ] [ Lokavidya ]

Spring 2015 : Graduated from University of Mumbai with Bachelors!!

Fun


Outside of work, I enjoy hiking, climbing, improv comedy, volunteering, competitive programming, hacking stuff over the weekend and contributing to open source .