Submitted by Dominik Hirner.

Submission data

Full name	fully-convolutional densely-connected neural network
Description	We propose a novel lightweight network for stereo estimation. The method uses densely connected layer structures to learn expressive features without the need of fully-connected layers or 3D convolutions. This leads to a network structure with only 0.37M parameters while still having competitive results. The post-processing consists of filtering, a consistency check and hole filling. This paper has been accepted to the ICPR 2020 conference in Milan which will be held on the 10-15 January 2021. Therefore this work has not yet been presented
Parameters	\eta = 6 \times 10^{-6}
Programming language(s)	python3, pytorch
Hardware	GeForce 2080 GTX
Source code or download URL	https://github.com/thedodo/fc-dcnn2
Submission creation date	25 Mar, 2020
Last edited	9 Nov, 2020

High-res multi-view results

Low-res many-view results

Low-res two-view results

SLAM results