Work collaboratively with our computer vision team to develop state of the art tools for ground truth data to train and validate our product computer vision stack. Our team has to solve all the computer vision product problems Magic Leap faces by owning our superpowers, having fearless curiosity, and finding the magic in the details. But, we can use whatever tools we can buy or build to get the job done. We use cutting edge technology in motion capture and 3D sensing (LIDAR etc.), as well as building our own ground truth tools using state of the art tools like these. This role would also work closely with others building solutions to the same sort of ground truth problems using robots and synthetic 3D photorealistic rendered data.
Build tools to create ground truth datasets using whatever means possible. Thinking outside the box in terms of software / electrical / mechanical / labeling / off-the-shelf systems to create ground truth systems as rapidly as possible.
Validate the precision / accuracy of created systems so data is of known quality.
Work in an interdisciplinary rapid prototyping environment to quickly build tools for dataset creation for head pose tracking, world reconstruction meshing, eye tracking, hand tracking, etc.
Scale tools with Data Operations team for large data collection
Understand deeply underlying technology to build state of the art ground truth tools
Must have 4+ years of relevant experience with commensurate educational degrees
Strong python and/or C++ programming skills
Experience building production quality solutions
Experience working with large sources of numerical data and solving problems with compute cloud
MS or PhD in CS, ECE, Physics or related fields
All your information will be kept confidential according to Equal Employment Opportunities guidelines.