About

Utkarsh Patel is a Computer Vision Engineer at Torc Robotics, specialized in machine learning, computer vision, and robotics. He has a solid foundation in ML, computer vision, software engineering, and robotics, having worked in a variety of internships and research labs throughout his undergraduate and graduate school. Currently, Utkarsh is developing real-time algorithms to solve one of the most challenging computer vision problems in the self-driving industry. Utkarsh has worked as a Software Engineer 2 - Autonomy at Torc Robotics, where he is involved in making self-driving Daimler trucks as part of the computer vision team. He has also worked as a Software Engineer Intern - Computer Vision at Torc Robotics, where he developed active learning algorithms to select the most uncertain data for the object detection model. He also developed techniques to calibrate neural network outputs so that the network's confidence scores are matched with the network's accuracy. Utkarsh also converted PyTorch models to TensorRT models. Utkarsh has worked as a Robotics Research Assistant at the University of Southern California, where he worked at the Robotic Embedded Systems Laboratory (RESL). He developed an opensource deep reinforcement learning (RL) framework named garage and designed and built an object detection/manipulation system for Sawyer robot. He also created Docker containers to automate training/testing of RL algorithms in simulation and with a real robot. Utkarsh has a Master of Science - MS degree in Computer Science (Intelligent Robotics) from the University of Southern California and a Bachelor of Science (B.Sc.) degree in Computer Science from Cleveland State University. His tech stack includes computer vision, software engineering, ML, object detection, research scientist, neural networks, PyTorch, deep learning, and search. With 3.19 years of relevant experience, Utkarsh is a highly skilled and experienced professional in computer vision and robotics.