Method

Fully-Convolutional Densely Connected Neural Network [FC-DCNN]
https://github.com/thedodo/fc-dcnn2

Submitted on 9 Nov. 2020 11:33 by
Dominik Hirner (ICG University of Technology Graz)

Running time:5 s
Environment:GPU @ >3.5 Ghz (Python)

Method 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}
Latex Bibtex:

Detailed Results

This page provides detailed results for the method(s) selected. For the first 20 test images, the percentage of erroneous pixels is depicted in the table. We use the error metric described in Object Scene Flow for Autonomous Vehicles (CVPR 2015), which considers a pixel to be correctly estimated if the disparity or flow end-point error is <3px or <5% (for scene flow this criterion needs to be fulfilled for both disparity maps and the flow map). Underneath, the left input image, the estimated results and the error maps are shown (for disp_0/disp_1/flow/scene_flow, respectively). The error map uses the log-color scale described in Object Scene Flow for Autonomous Vehicles (CVPR 2015), depicting correct estimates (<3px or <5% error) in blue and wrong estimates in red color tones. Dark regions in the error images denote the occluded pixels which fall outside the image boundaries. The false color maps of the results are scaled to the largest ground truth disparity values / flow magnitudes.

Test Set Average

Error D1-bg D1-fg D1-all
All / All 5.21 15.16 6.87
All / Est 5.21 15.16 6.87
Noc / All 4.74 13.95 6.26
Noc / Est 4.74 13.95 6.26
This table as LaTeX

Test Image 0

Error D1-bg D1-fg D1-all
All / All 6.81 1.96 6.15
All / Est 6.81 1.96 6.15
Noc / All 6.76 1.96 6.09
Noc / Est 6.76 1.96 6.09
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 1

Error D1-bg D1-fg D1-all
All / All 4.63 8.16 5.02
All / Est 4.63 8.16 5.02
Noc / All 4.33 8.16 4.77
Noc / Est 4.33 8.16 4.77
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 2

Error D1-bg D1-fg D1-all
All / All 5.76 5.50 5.75
All / Est 5.76 5.50 5.75
Noc / All 5.07 5.50 5.09
Noc / Est 5.07 5.50 5.09
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 3

Error D1-bg D1-fg D1-all
All / All 6.10 10.88 6.54
All / Est 6.10 10.88 6.54
Noc / All 5.68 10.88 6.16
Noc / Est 5.68 10.88 6.16
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 4

Error D1-bg D1-fg D1-all
All / All 7.04 10.86 7.68
All / Est 7.04 10.86 7.68
Noc / All 5.99 10.86 6.81
Noc / Est 5.99 10.86 6.81
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 5

Error D1-bg D1-fg D1-all
All / All 12.87 11.78 12.78
All / Est 12.87 11.78 12.78
Noc / All 11.52 11.78 11.54
Noc / Est 11.52 11.78 11.54
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 6

Error D1-bg D1-fg D1-all
All / All 14.33 5.59 13.41
All / Est 14.33 5.59 13.41
Noc / All 13.70 5.59 12.83
Noc / Est 13.70 5.59 12.83
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 7

Error D1-bg D1-fg D1-all
All / All 2.87 6.33 3.55
All / Est 2.87 6.33 3.55
Noc / All 2.78 6.33 3.49
Noc / Est 2.78 6.33 3.49
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 8

Error D1-bg D1-fg D1-all
All / All 2.71 2.94 2.75
All / Est 2.71 2.94 2.75
Noc / All 2.69 2.94 2.74
Noc / Est 2.69 2.94 2.74
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 9

Error D1-bg D1-fg D1-all
All / All 2.79 2.47 2.71
All / Est 2.79 2.47 2.71
Noc / All 2.78 2.42 2.69
Noc / Est 2.78 2.42 2.69
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 10

Error D1-bg D1-fg D1-all
All / All 2.74 5.61 3.39
All / Est 2.74 5.61 3.39
Noc / All 2.71 5.61 3.38
Noc / Est 2.71 5.61 3.38
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 11

Error D1-bg D1-fg D1-all
All / All 2.72 1.84 2.56
All / Est 2.72 1.84 2.56
Noc / All 2.55 1.84 2.42
Noc / Est 2.55 1.84 2.42
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 12

Error D1-bg D1-fg D1-all
All / All 1.63 2.93 1.71
All / Est 1.63 2.93 1.71
Noc / All 1.55 2.93 1.64
Noc / Est 1.55 2.93 1.64
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 13

Error D1-bg D1-fg D1-all
All / All 1.48 3.02 1.67
All / Est 1.48 3.02 1.67
Noc / All 1.31 3.02 1.52
Noc / Est 1.31 3.02 1.52
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 14

Error D1-bg D1-fg D1-all
All / All 2.71 5.34 2.76
All / Est 2.71 5.34 2.76
Noc / All 2.54 5.34 2.59
Noc / Est 2.54 5.34 2.59
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 15

Error D1-bg D1-fg D1-all
All / All 4.52 3.69 4.45
All / Est 4.52 3.69 4.45
Noc / All 4.47 3.69 4.39
Noc / Est 4.47 3.69 4.39
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 16

Error D1-bg D1-fg D1-all
All / All 8.05 3.75 7.42
All / Est 8.05 3.75 7.42
Noc / All 7.42 3.75 6.87
Noc / Est 7.42 3.75 6.87
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 17

Error D1-bg D1-fg D1-all
All / All 3.26 2.62 3.19
All / Est 3.26 2.62 3.19
Noc / All 2.95 2.62 2.91
Noc / Est 2.95 2.62 2.91
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 18

Error D1-bg D1-fg D1-all
All / All 9.06 14.13 11.47
All / Est 9.06 14.13 11.47
Noc / All 8.50 14.13 11.20
Noc / Est 8.50 14.13 11.20
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 19

Error D1-bg D1-fg D1-all
All / All 2.18 4.68 2.46
All / Est 2.18 4.68 2.46
Noc / All 2.06 4.68 2.36
Noc / Est 2.06 4.68 2.36
This table as LaTeX

Input Image

D1 Result

D1 Error




eXTReMe Tracker