Unsupervised Learning of Dense Optical Flow, Depth and Egomotion from Sparse Event Data [Updated]

We present unsupervised learning of depth and motion from sparse event data generated by a Dynamic Vision Sensor (DVS). Our work is the first that generates dense depth and optical flow information from sparse event data. Results show significant improvements upon previous works that used deep learning for flow estimation from both images and events. The preprint is available on the arXiv: https://arxiv.org/abs/1809.08625