About

Kratarth Goel is an experienced Research Scientist with a strong background in Deep Learning, Computer Vision, Spatio-Temporal Analysis, and Trajectory prediction. He is currently working as a Staff Research Scientist at Waymo, where he is involved in end-to-end perception and prediction. Prior to this, he worked at Zoox as a Senior Research Engineer/TLM (Computer Vision/Perception) and Research Engineer (Deep Learning and Computer Vision). At Zoox, he was responsible for the entire vision semantic scene understanding stack, including Detection, Tracking, Panoptic Segmentation, Depth Estimation, 3D Object Detection, Semantic Attributes, Long-Tail Reasoning, Detecting World Change, and more. He led and delivered multiple solutions end-to-end, from problem conceptions to solution proposals, implementation, and research, to machine learning acceleration for on-vehicle showtime. He improved the performance of the company's top-line metrics for the perception stack many folds on increasingly challenging metrics continuously. He also helped build and grow a team of dedicated engineers delivering state-of-the-art solutions. He was intricately involved with and cross-organization projects to set direction for the entire perception stack, including ML accelerations, ML infrastructure, advanced sensor engineering, prediction, perception risk analysis (metrics and validations pipeline), data selection, and annotation. He has published work on QuadroNet: Multi-Task Learning for Semantic Depth Aware Instance Segmentation and has 12 US granted patents (multiple pending) on self-driving technology. Kratarth has a Master’s Degree in Computer Science from Stanford University and a B.E. Hons in Computer Science from Birla Institute of Technology and Science. He has a strong research background and is highly skilled in his field. He is proficient in various technologies, including Research Scientist, Computer Vision, Deep Learning, ML, Object detection, Object Detection, and infra. With 5.87 years of relevant experience, Kratarth is a strong research professional who has demonstrated his expertise in the autonomous vehicles industry.