Method

Channel affinity enhancing feature network for stereo matching [CAEF-Net]
[Anonymous Submission]

Submitted on 27 Nov. 2020 08:46 by
[Anonymous Submission]

Running time:0.44 s
Environment:1 core @ 2.5 Ghz (Python)

Method Description:
首先本文提出的网络设计了亲和力增强模块,利用注意力
机制学习不同通道之间内在的亲和力,并利用它来重构代
价特征,重构后的代价特征增强了具有有效信息的通道。
其次,开发了多尺度融合模块,通过融合图像中的多尺度
语义信息和多尺度视差获得精细视差,进一步提升性能。
最后,设计了一种加权损失函数,它能够根据视差误差计
算自适应权重引导网络从像素级视差学习逐渐过渡到亚像
素级,提升训练效果。
Parameters:
\alpha=0.2
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 1.68 3.92 2.05
All / Est 1.68 3.92 2.05
Noc / All 1.53 3.57 1.87
Noc / Est 1.53 3.57 1.87
This table as LaTeX

Test Image 0

Error D1-bg D1-fg D1-all
All / All 1.48 0.66 1.36
All / Est 1.48 0.66 1.36
Noc / All 1.42 0.66 1.31
Noc / Est 1.42 0.66 1.31
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 1

Error D1-bg D1-fg D1-all
All / All 1.74 1.58 1.72
All / Est 1.74 1.58 1.72
Noc / All 1.63 1.58 1.63
Noc / Est 1.63 1.58 1.63
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 2

Error D1-bg D1-fg D1-all
All / All 2.29 1.49 2.25
All / Est 2.29 1.49 2.25
Noc / All 2.11 1.49 2.08
Noc / Est 2.11 1.49 2.08
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 3

Error D1-bg D1-fg D1-all
All / All 2.09 0.29 1.92
All / Est 2.09 0.29 1.92
Noc / All 2.03 0.29 1.87
Noc / Est 2.03 0.29 1.87
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 4

Error D1-bg D1-fg D1-all
All / All 3.07 0.39 2.62
All / Est 3.07 0.39 2.62
Noc / All 3.13 0.39 2.67
Noc / Est 3.13 0.39 2.67
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 5

Error D1-bg D1-fg D1-all
All / All 3.68 0.64 3.41
All / Est 3.68 0.64 3.41
Noc / All 3.40 0.64 3.15
Noc / Est 3.40 0.64 3.15
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 6

Error D1-bg D1-fg D1-all
All / All 3.75 0.62 3.42
All / Est 3.75 0.62 3.42
Noc / All 3.83 0.62 3.49
Noc / Est 3.83 0.62 3.49
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 7

Error D1-bg D1-fg D1-all
All / All 0.55 2.17 0.87
All / Est 0.55 2.17 0.87
Noc / All 0.56 2.17 0.88
Noc / Est 0.56 2.17 0.88
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 8

Error D1-bg D1-fg D1-all
All / All 0.51 1.01 0.60
All / Est 0.51 1.01 0.60
Noc / All 0.51 1.01 0.60
Noc / Est 0.51 1.01 0.60
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 9

Error D1-bg D1-fg D1-all
All / All 0.58 1.27 0.76
All / Est 0.58 1.27 0.76
Noc / All 0.59 1.33 0.77
Noc / Est 0.59 1.33 0.77
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 10

Error D1-bg D1-fg D1-all
All / All 1.01 1.95 1.22
All / Est 1.01 1.95 1.22
Noc / All 1.01 1.95 1.22
Noc / Est 1.01 1.95 1.22
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 11

Error D1-bg D1-fg D1-all
All / All 1.06 0.61 0.98
All / Est 1.06 0.61 0.98
Noc / All 1.06 0.61 0.98
Noc / Est 1.06 0.61 0.98
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 12

Error D1-bg D1-fg D1-all
All / All 0.54 0.37 0.53
All / Est 0.54 0.37 0.53
Noc / All 0.42 0.37 0.42
Noc / Est 0.42 0.37 0.42
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 13

Error D1-bg D1-fg D1-all
All / All 0.82 0.04 0.72
All / Est 0.82 0.04 0.72
Noc / All 0.72 0.04 0.63
Noc / Est 0.72 0.04 0.63
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 14

Error D1-bg D1-fg D1-all
All / All 1.40 0.06 1.37
All / Est 1.40 0.06 1.37
Noc / All 1.31 0.06 1.29
Noc / Est 1.31 0.06 1.29
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 15

Error D1-bg D1-fg D1-all
All / All 3.14 0.19 2.87
All / Est 3.14 0.19 2.87
Noc / All 3.20 0.19 2.93
Noc / Est 3.20 0.19 2.93
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 16

Error D1-bg D1-fg D1-all
All / All 3.90 0.10 3.34
All / Est 3.90 0.10 3.34
Noc / All 3.73 0.10 3.19
Noc / Est 3.73 0.10 3.19
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 17

Error D1-bg D1-fg D1-all
All / All 1.16 0.01 1.04
All / Est 1.16 0.01 1.04
Noc / All 1.13 0.01 1.01
Noc / Est 1.13 0.01 1.01
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 18

Error D1-bg D1-fg D1-all
All / All 5.16 0.84 3.11
All / Est 5.16 0.84 3.11
Noc / All 5.13 0.84 3.07
Noc / Est 5.13 0.84 3.07
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 19

Error D1-bg D1-fg D1-all
All / All 0.99 0.58 0.94
All / Est 0.99 0.58 0.94
Noc / All 0.96 0.58 0.92
Noc / Est 0.96 0.58 0.92
This table as LaTeX

Input Image

D1 Result

D1 Error




eXTReMe Tracker