Dense Depth Posterior (DDP) From Single Image and Sparse Range

Yanchao Yang, Alex Wong, Stefano Soatto; Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2019, pp. 3353-3362

Abstract


We present a deep learning system to infer the posterior distribution of a dense depth map associated with an image, by exploiting sparse range measurements, for instance from a lidar. While the lidar may provide a depth value for a small percentage of the pixels, we exploit regularities reflected in the training set to complete the map so as to have a probability over depth for each pixel in the image. We exploit a Conditional Prior Network, that allows associating a probability to each depth value given an image, and combine it with a likelihood term that uses the sparse measurements. Optionally we can also exploit the availability of stereo during training, but in any case only require a single image and a sparse point cloud at run-time. We test our approach on both unsupervised and supervised depth completion using the KITTI benchmark, and improve the state-of-the-art in both.

Related Material


[pdf]
[bibtex]
@InProceedings{Yang_2019_CVPR,
author = {Yang, Yanchao and Wong, Alex and Soatto, Stefano},
title = {Dense Depth Posterior (DDP) From Single Image and Sparse Range},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {June},
year = {2019}
}