Vishaal Udandarao
Description: MSR, CMU.
jekyll (1644) deep learning (1274) computer vision (868) neural networks (239) cmu (159) yoda (30) iiit (17) sdslabs (4) sarthak268 (1) sarthak bhagat (1)
I am a graduate student at Robotics Institute , Carnegie Mellon University . I am working at Advanced Agents - Robotics Technology Lab advised by Dr. Katia Sycara . Currently, I am exploring the following problems: (1) manipulation of novel objects using LLMs via affordance reasoning, (2) short-context action anticipation for human-robot collaboration in manipulation tasks, (3) learning of visual (objects and scenes) and abstract (affordances and attributes) concepts using domain knowledge priors.
Before joining CMU, I was a Deep Learning Research Engineer at Preimage , where I specialized in problems involving 3D vision for drone-based photogrammetry. Previously, I worked as a Visiting Researcher at Cognitive Learning and Vision for Robotics (CLVR) Lab , University of Southern California (USC), USA under the guidance of Dr. Joseph Lim . My work there centered on the adaptation of policies to novel unseen environments and continual learning of skills. During my short stint as a Research Assistant at
I graduated with my B.Tech (with Honors) from IIIT-Delhi . My Bachelor’s Thesis, completed under Dr. Saket Anand in collaboration with Dr. Pavan Turaga, Geometric Media Lab (Arizona State University, USA) at the Infosys Center of Artificial Intelligence , involved unsupervised representation learning to disentangle multiple factors of variation in images. You can find my Bachelor’s thesis here . I also interned at the SUTD Brain Lab under Dr. Nengli Lim , where I worked on the disentanglement of video seque
Vishaal Udandarao