English | EN Take a bunch of Google Street View photo’s and apply deep learning. The outcome is “a video” and something that’s called Deep Stereo. It’s and end-to-end trained system that fills in the gaps between the pictures and make you feel as if you can walk in the scene like 3D. Deep networks have recently enjoyed enormous success when applied to recognition and classification problems in computer vision, but their use in graphics problems has been limited.
DeepStereo: Learning to Predict New Views from the World’s Imagery
About the author
Menno is Director of the Sogeti Research Institute for the Analysis of New Technology (VINT). He mixes personal life experiences with the findings of the 19 years of research done at the VINT Research Institute. Menno has co-authored many books on the impact of new technology on business and society.