Soham Kulkarni
Second-year MS student in Computer Science at UCLA.
Los Angeles, CA
Email: [click to reveal]
GitHub: stochasticritic
LinkedIn: Soham Kulkarni
“Stochasticritic—life’s randomness (stochastic) unfolds, while the critic (ego) judges, veiling the true Self—until awareness dissolves all illusions.”
Background & Research
Hi!, I'm Soham. I did my undergrad at IIT Hyderabad in Electrical Engineering with a minor in Entrepreneurship, and I'm presently pursuing my Master's in CS at UCLA, researching in Robot Learning.
My background lies in Robotics, particularly in Perception and Robot Learning. During my undergraduate studies at IIT Hyderabad, I worked on developing autonomous mobile robots (hardware + low-level controls) for data acquisition and surveillance, as well as using various RL algorithms for learning navigation policies for these robots in simulation (under Dr. M Vidyasagar FRS).
During my graduate studies at UCLA, I developed a close-range planning and navigation suite for the ARTEMIS humanoid robot (that won the Robocup '24 humanoids league) that is critical when the main vision pipeline fails or goes through a blackout (under Dr. Dennis Hong, RoMeLa). Currently, I'm exploring effective action representations for imitation learning and evaluating task-specific trajectories for robot manipulation tasks (under Dr. Yuchen Cui, UCLA Robot Intelligence Lab).
Hobbies & Interests
In my free time, I love going on nature expeditions for wildlife photography, reading philosophy, painting, running, or hitting the gym. Lately, I've been training in various martial arts (Muay Thai, Brazilian Jiu Jitsu, Kali, and Boxing), and I've recently picked up an interest in tennis.