Nikhil Nakhate

I'm

About

Roboticist/Deep Learning researcher

Autonomous Robots | Computer Vision | Deep Learning | Structured Light

I am Machine Learning researcher working on multi-cam multi-task learning models. I am a passionate roboticist on a quest to find newer and better ways to build robots that will help humans automate dangerous, mundane and tedius tasks. I currently work in the autonomous driving perception team at Lucid motors. I have previously worked at Tesla autopilot and a robot barista startup, Nourish.

Resume

Education

Master of Computer Science

2016 - 2018

University of Wisconsin-Madison, Madison, WI

Specialized in computer vision

Bachelor of Technology in Computer Science

2008 - 2012

National Institute of Technology, Calicut, KL, India

Professional Experience

Technical Specialist (Staff Engineer), Perception Algorithms

December 2020 - Present

Lucid Motors, Bay Area, CA

  • Project lead: Multi-task learning model framework, 3D lane detection
  • Prototyped and deployed various monocular 3D object detection models
  • Lane classification for multiple lane detection models
  • Multi-agent trajectory prediction using VAE and Transformer architecture
  • Tooling: Developed tools for data analysis and metrics

Robotics Software Engineer

December 2019 - October 2020

Nourish Technology, Bay Area, CA

  • Object detection/anomaly prediction for gantry robot
  • Environment mapping for a robot arm in a distributed robotic system
  • Work on a simulator project using Unity/Gazebo (ROS)

Projects

Structured Light

Structured Light is an active stereo method for 3D reconstruction. In this work we use space filling curves like Hilbert and Hamiltonian curves and show that Hamiltonian curves give the best result and are most robust to various ambient light conditions

Udacity Nanodegree

Completed the Udacity nanodegree and learned about perception and control

BBox Cafe

BBox Cafe was a coffee shop operated by a robot barista. I worked on the computer vision for object detection, environment mapping and robot arm manipulation

Music