Temporally Consistent Depth Estimation in Videos with Recurrent Architectures

Denis Tananaev, Huizhong Zhou, Benjamin Ummenhofer, Thomas Brox; Proceedings of the European Conference on Computer Vision (ECCV) Workshops, 2018, pp. 0-0

Abstract


Convolutional networks trained on large RGB-D datasets have enabled depth estimation from a single image. Many works on automotive applications rely on such approaches. However, all existing methods work on a frame-by-frame manner when applied to videos, which leads to inconsistent depth estimates over time. In this paper, we introduce for the first time an approach that yields temporally consistent depth estimates over multiple frames of a video. This is done by a dedicated architecture based on convolutional LSTM units and layer normalization. Our approach achieves superior performance on several error metrics when compared to independent frame processing. This also shows in an improved quality of the reconstructed multi-view point clouds.

Related Material


[pdf]
[bibtex]
@InProceedings{Tananaev_2018_ECCV_Workshops,
author = {Tananaev, Denis and Zhou, Huizhong and Ummenhofer, Benjamin and Brox, Thomas},
title = {Temporally Consistent Depth Estimation in Videos with Recurrent Architectures},
booktitle = {Proceedings of the European Conference on Computer Vision (ECCV) Workshops},
month = {September},
year = {2018}
}