Deep learning algorithms develop spatial awareness - converting 2D images and video into estimates of 3D depths. These estimations of how far objects are can be streamed and processed in real time from a camera to a convolutional neural network. This clip demonstrates the current state of the art for Computer Vision algorithms aiming to achieve real time depth detection.
Neural network input is a RGB image and the output is a relative depth map where the range is the closest to farthest points in the scene. The network was trained on the NYU v2 depth data set.
To learn more about this class of algorithms or for tutorials on how to set up your own, visit: [ Ссылка ]
Videographers: Oleg Magni, Mike, Taryn Elliott, Anthony, Kelly Lacy
Music: Composed by a deep learning algorithm. For more details visit [ Ссылка ]
