Method

Selective Attention-based Residual Network [at] [SARNet]


Submitted on 17 Apr. 2026 09:34 by
Bofa Liang (Sun Yat-sen University)

Running time:0.01 s
Environment:GPU @ 2.5 Ghz (Python)

Method Description:
SARNet is a real-time, high-precision stereo
matching network tailored for resource-constrained
edge devices. We propose a Selective Attention
Feature Extractor (SAFE) to extract compact multi-
scale features, and construct the cost volume
using a compute-efficient Group-wise L2 distance.
Cost aggregation is performed by an Enhanced
Residual Cost Aggregation (ERCA) module via
shallow 3D convolutions and hybrid normalization,
followed by a Confidence-weighted Residual
Refinement (CWRR) module that suppresses erroneous
updates in low-confidence regions. Note: This work
is currently under review in IEEE Transactions on
Vehicular Technology (TVT).
Parameters:
See paper
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 1.87 4.45 2.30
All / Est 1.87 4.45 2.30
Noc / All 1.69 4.09 2.09
Noc / Est 1.69 4.09 2.09
This table as LaTeX

Test Image 0

Error D1-bg D1-fg D1-all
All / All 1.90 0.39 1.69
All / Est 1.90 0.39 1.69
Noc / All 1.85 0.39 1.65
Noc / Est 1.85 0.39 1.65
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 1

Error D1-bg D1-fg D1-all
All / All 1.68 2.20 1.74
All / Est 1.68 2.20 1.74
Noc / All 1.56 2.20 1.63
Noc / Est 1.56 2.20 1.63
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 2

Error D1-bg D1-fg D1-all
All / All 2.56 1.11 2.48
All / Est 2.56 1.11 2.48
Noc / All 2.23 1.11 2.18
Noc / Est 2.23 1.11 2.18
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 3

Error D1-bg D1-fg D1-all
All / All 2.31 0.28 2.12
All / Est 2.31 0.28 2.12
Noc / All 2.27 0.28 2.08
Noc / Est 2.27 0.28 2.08
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 4

Error D1-bg D1-fg D1-all
All / All 3.05 2.41 2.94
All / Est 3.05 2.41 2.94
Noc / All 3.05 2.41 2.94
Noc / Est 3.05 2.41 2.94
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 5

Error D1-bg D1-fg D1-all
All / All 3.08 0.81 2.87
All / Est 3.08 0.81 2.87
Noc / All 2.58 0.81 2.41
Noc / Est 2.58 0.81 2.41
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 6

Error D1-bg D1-fg D1-all
All / All 4.93 1.50 4.56
All / Est 4.93 1.50 4.56
Noc / All 5.03 1.50 4.65
Noc / Est 5.03 1.50 4.65
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 7

Error D1-bg D1-fg D1-all
All / All 0.31 3.09 0.85
All / Est 0.31 3.09 0.85
Noc / All 0.31 3.09 0.86
Noc / Est 0.31 3.09 0.86
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 8

Error D1-bg D1-fg D1-all
All / All 0.38 2.17 0.71
All / Est 0.38 2.17 0.71
Noc / All 0.37 2.17 0.70
Noc / Est 0.37 2.17 0.70
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 9

Error D1-bg D1-fg D1-all
All / All 0.35 1.41 0.62
All / Est 0.35 1.41 0.62
Noc / All 0.35 1.48 0.63
Noc / Est 0.35 1.48 0.63
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 10

Error D1-bg D1-fg D1-all
All / All 1.07 2.42 1.38
All / Est 1.07 2.42 1.38
Noc / All 1.08 2.42 1.39
Noc / Est 1.08 2.42 1.39
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 11

Error D1-bg D1-fg D1-all
All / All 0.88 0.60 0.83
All / Est 0.88 0.60 0.83
Noc / All 0.89 0.60 0.83
Noc / Est 0.89 0.60 0.83
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 12

Error D1-bg D1-fg D1-all
All / All 0.72 0.48 0.71
All / Est 0.72 0.48 0.71
Noc / All 0.59 0.48 0.58
Noc / Est 0.59 0.48 0.58
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 13

Error D1-bg D1-fg D1-all
All / All 0.75 0.06 0.67
All / Est 0.75 0.06 0.67
Noc / All 0.71 0.06 0.63
Noc / Est 0.71 0.06 0.63
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 14

Error D1-bg D1-fg D1-all
All / All 1.30 0.00 1.27
All / Est 1.30 0.00 1.27
Noc / All 1.16 0.00 1.14
Noc / Est 1.16 0.00 1.14
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 15

Error D1-bg D1-fg D1-all
All / All 2.43 0.61 2.26
All / Est 2.43 0.61 2.26
Noc / All 2.48 0.61 2.30
Noc / Est 2.48 0.61 2.30
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 16

Error D1-bg D1-fg D1-all
All / All 3.55 0.14 3.05
All / Est 3.55 0.14 3.05
Noc / All 3.33 0.14 2.86
Noc / Est 3.33 0.14 2.86
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 17

Error D1-bg D1-fg D1-all
All / All 0.83 0.11 0.75
All / Est 0.83 0.11 0.75
Noc / All 0.81 0.11 0.73
Noc / Est 0.81 0.11 0.73
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 18

Error D1-bg D1-fg D1-all
All / All 4.68 1.28 3.06
All / Est 4.68 1.28 3.06
Noc / All 4.52 1.28 2.97
Noc / Est 4.52 1.28 2.97
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 19

Error D1-bg D1-fg D1-all
All / All 0.88 0.58 0.85
All / Est 0.88 0.58 0.85
Noc / All 0.86 0.58 0.83
Noc / Est 0.86 0.58 0.83
This table as LaTeX

Input Image

D1 Result

D1 Error




eXTReMe Tracker