MotionCorrect, Inc. consists of a small group of passionate applied computer vision specialists based in Montreal interested in solving real world problems related to teaching robots and computers to see the world smarter.
Our current main project involves developing an optical tracking solution to spatially monitor position in real time. Application of our solution includes motion monitoring in MRI environment, spatial position tracking in augmented reality situations, pose tracking during motion capturing.
Our latest effort in incorporating deep learning (Project Code Name: PoseTracker) to solve part of this problem has won the 2nd place in 2018 Fall Microsoft AI Lab Idea Challenge as web-archived here. You can learn more about PoseTracker, the project here. That particular project would not have been possible if not for the various backend traditional computer vision projects that paved the way to get us there.
We also have wide ranging interests in other applied computer vision topics such as regional temperature data prediction via super resolution upsampling using convolutional neural network, help optimizing architecture and training of quality control via computer vision neural network for semiconductor manufacturers, or help robots find targeted optical stripes for target and position alignment used in First Robotics Competition.
Should you require our expertise or simply would like to have a nice chat about the future and perspective of computer vision and where it can make real world impact (beyond the incessant AI CNN hype), feel free to reach out to us. We don’t bite.