SBX Robotics

Synthetic data for better vision.

Y CombinatorWinter 2021InactiveB2BEngineering, Product and DesignArtificial IntelligenceMachine LearningRoboticsComputer Vision

What they do

Problem

Robotics teams lack large amounts of accurately labeled vision training data, making it slow and expensive to train computer vision models.

Solution

It generates perfectly labeled synthetic robot-camera training images using simulation software from a small set of real images.

Who it's for

Robotics companies and teams training deep learning vision models for robots.

Founders

Ian Dewancker

Founder

Led applied research projects at Uber ATG, Kindred AI, and SigOpt focusing on computer vision, robotics, and optimization for machine learning. MSc in CS from UBC (2013), Engineering at Waterloo (2011).

LinkedIn · X

Artem Avdacev

Founder

Building something new in the intersection of computer vision, robotics, and computer graphics (games & VFX) Before SBX, I spent 7 years at Yelp fighting fraud and fake reviews as an engineer & PM, and then took off to sail across the pacific ocean, dive with sharks, and hike the himalayas. Based in Montreal, Canada.