SLAMcore uses deep learning to help robots, consumer products and drones identify objects and people around them, to work effectively in dynamic environments. Deep learning semantics enhances localization and enhances maps for more accurate path planning, obstacle avoidance - even in dynamic real-world environments. Factory robots, for example, can differentiate between static equipment, pallets, human workers and vehicles, allowing it to avoid dangerous areas and accurately navigate
RGB data is run through a neural network and overlaid on a point-cloud for 3D understanding
SLAMCore's algorithms remove dynamic objects from the scene, increasing mapping robustness
SLAMCore's algorithms can better estimate depth from a camera's video stream