Aug 2021 – Dec 2025 · 4 yrs 4 mos · San Mateo, California, United States

Jan 2021 – May 2021 · 4 mos · Champaign, Illinois, United States

• - Self supervised learning on egocentric videos to learn about objects and their states

Aug 2020 – Dec 2020 · 4 mos · Champaign, Illinois, United States

• TA for the course ECE598: Learning-Based Robotics

Aug 2019 – May 2020 · 9 mos · Urbana-Champaign, Illinois Area

• Formulated a data-driven strategy to overcome the shortcomings of heuristic based active learning strategies
• Designed a neural architecture to rank a set of unordered points
• Conducted experiments and proposed new metrics to evaluate the efficacy of the algorithm

May 2020 – Aug 2020 · 3 mos · San Francisco Bay Area

• Formulated and implemented projection, voxel and point-based models for semantically segmenting large point clouds

Jul 2018 – Aug 2019 · 1 yr 1 mo · New Delhi Area, India

• Developed conversational agents for goal-oriented dialog, using deep learning and reinforcement learning techniques
• Tackled commonly faced problems both in natural language understanding and language generation tasks
• Employed machine learning techniques to help automate and boost the efficiency of technical support

Jan 2018 – Apr 2018 · 3 mos · Kanpur Area, India

• Tutor for the course Fundamentals of Computing (ESC101)
• Responsible for conducting weekly tutorial sessions, setting exams and assignments, and ensuring smooth conduct of lab sessions.

May 2017 – Jul 2017 · 2 mos · Amherst, Massachusetts

• Collaboratively devised an algorithm that, given a 2D or 3D shape as input, could accurately predict programs to reconstruct given shape, based on principles of Constructive Solid Geometry (CSG)
• Designed a grammar to represent visual programs and developed an accompanying renderer to parse programs and produce corresponding outputs
• Curated artificial dataset of image-program pairs to train a recurrent neural network for visual program induction