Method

Dual-Matching Net [DuMa-Net]


Submitted on 25 Jun. 2022 13:29 by
Shuqiao Sun (BUAA)

Running time:0.38 s
Environment:PyTorch GPU

Method Description:
In this paper, stereo matching is reconsidered
using a specially designed dual-matching method
with cross-attention mechanism which liberates the
algorithm from manually pre-specified disparity
ranges and the performance is guaranteed without
re-training when the camera rig varies. Moreover,
to tackle the mismatches on edges and details, an
exquisite module is designed based on left-right
consistency which further refines the estimated
disparity map.
Parameters:
weights are [1, 0.5, 0.5, 1, 0.5, 0.5]
alpha = 1
beta = 0.1
Latex Bibtex:
@article{sunrange,
title={Range-free disparity estimation with self-
adaptive dual-matching},
author={Sun, Shuqiao and liu, Rongke and Sun,
Shantong},
journal={IET Computer Vision},
publisher={Wiley Online Library}
}

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 1.40 3.18 1.70
All / Est 1.40 3.18 1.70
Noc / All 1.26 2.84 1.52
Noc / Est 1.26 2.84 1.52
This table as LaTeX

Test Image 0

Error D1-bg D1-fg D1-all
All / All 1.18 0.88 1.14
All / Est 1.18 0.88 1.14
Noc / All 1.14 0.88 1.10
Noc / Est 1.14 0.88 1.10
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 1

Error D1-bg D1-fg D1-all
All / All 1.58 2.69 1.70
All / Est 1.58 2.69 1.70
Noc / All 1.51 2.69 1.64
Noc / Est 1.51 2.69 1.64
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 2

Error D1-bg D1-fg D1-all
All / All 2.09 1.43 2.06
All / Est 2.09 1.43 2.06
Noc / All 1.97 1.43 1.95
Noc / Est 1.97 1.43 1.95
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 3

Error D1-bg D1-fg D1-all
All / All 1.71 0.09 1.56
All / Est 1.71 0.09 1.56
Noc / All 1.65 0.09 1.50
Noc / Est 1.65 0.09 1.50
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 4

Error D1-bg D1-fg D1-all
All / All 0.59 0.29 0.54
All / Est 0.59 0.29 0.54
Noc / All 0.38 0.29 0.37
Noc / Est 0.38 0.29 0.37
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 5

Error D1-bg D1-fg D1-all
All / All 2.24 1.49 2.17
All / Est 2.24 1.49 2.17
Noc / All 1.92 1.49 1.88
Noc / Est 1.92 1.49 1.88
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 6

Error D1-bg D1-fg D1-all
All / All 4.04 1.16 3.74
All / Est 4.04 1.16 3.74
Noc / All 4.12 1.16 3.80
Noc / Est 4.12 1.16 3.80
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 7

Error D1-bg D1-fg D1-all
All / All 0.45 3.15 0.98
All / Est 0.45 3.15 0.98
Noc / All 0.45 3.15 0.99
Noc / Est 0.45 3.15 0.99
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 8

Error D1-bg D1-fg D1-all
All / All 0.39 1.62 0.62
All / Est 0.39 1.62 0.62
Noc / All 0.38 1.62 0.61
Noc / Est 0.38 1.62 0.61
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 9

Error D1-bg D1-fg D1-all
All / All 0.43 1.05 0.59
All / Est 0.43 1.05 0.59
Noc / All 0.44 1.10 0.60
Noc / Est 0.44 1.10 0.60
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 10

Error D1-bg D1-fg D1-all
All / All 1.23 2.45 1.51
All / Est 1.23 2.45 1.51
Noc / All 1.24 2.45 1.52
Noc / Est 1.24 2.45 1.52
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 11

Error D1-bg D1-fg D1-all
All / All 0.78 0.30 0.69
All / Est 0.78 0.30 0.69
Noc / All 0.78 0.30 0.70
Noc / Est 0.78 0.30 0.70
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 12

Error D1-bg D1-fg D1-all
All / All 0.61 0.56 0.61
All / Est 0.61 0.56 0.61
Noc / All 0.46 0.56 0.47
Noc / Est 0.46 0.56 0.47
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 13

Error D1-bg D1-fg D1-all
All / All 0.63 0.07 0.56
All / Est 0.63 0.07 0.56
Noc / All 0.50 0.07 0.45
Noc / Est 0.50 0.07 0.45
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 14

Error D1-bg D1-fg D1-all
All / All 1.26 0.00 1.23
All / Est 1.26 0.00 1.23
Noc / All 1.15 0.00 1.13
Noc / Est 1.15 0.00 1.13
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 15

Error D1-bg D1-fg D1-all
All / All 2.32 0.23 2.13
All / Est 2.32 0.23 2.13
Noc / All 2.37 0.23 2.17
Noc / Est 2.37 0.23 2.17
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 16

Error D1-bg D1-fg D1-all
All / All 3.35 0.18 2.88
All / Est 3.35 0.18 2.88
Noc / All 3.14 0.18 2.70
Noc / Est 3.14 0.18 2.70
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 17

Error D1-bg D1-fg D1-all
All / All 0.66 0.02 0.60
All / Est 0.66 0.02 0.60
Noc / All 0.65 0.02 0.59
Noc / Est 0.65 0.02 0.59
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 18

Error D1-bg D1-fg D1-all
All / All 4.28 1.62 3.01
All / Est 4.28 1.62 3.01
Noc / All 4.14 1.62 2.93
Noc / Est 4.14 1.62 2.93
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 19

Error D1-bg D1-fg D1-all
All / All 0.72 0.64 0.71
All / Est 0.72 0.64 0.71
Noc / All 0.69 0.64 0.69
Noc / Est 0.69 0.64 0.69
This table as LaTeX

Input Image

D1 Result

D1 Error




eXTReMe Tracker