Method

LightStereo-S [LightStereo-S]
https://github.com/XiandaGuo/OpenStereo

Submitted on 4 Jul. 2024 06:07 by
Yiqun Duan (University of Technology Sydney)

Running time:0.01 s
Environment:1 core @ 2.5 Ghz (Python)

Method Description:
We present LightStereo, a cutting-edge stereo-
matching network crafted to accelerate the matching
process. Departing from conventional methodologies
that rely on aggregating computationally intensive
4D costs, LightStereo adopts the 3D cost volume as a
lightweight alternative. While similar approaches
have been explored previously, our breakthrough lies
in enhancing performance through a dedicated focus
on the channel dimension of the 3D cost volume,
where the distribution of matching costs is
encapsulated. Our exhaustive exploration has yielded
plenty of strategies to amplify the capacity of the
pivotal dimension, ensuring both precision and
efficiency. We compare the proposed LightStereo with
existing state-of-the-art methods across various
benchmarks, which demonstrate its superior
performance in speed, accuracy, and resource
utilization.
Parameters:
3.44
Latex Bibtex:
@inproceedings{guo2025lightstereo,
title={Lightstereo: Channel boost is all you
need for efficient 2d cost aggregation},
author={Guo, Xianda and Zhang, Chenming and
Zhang, Youmin and Zheng, Wenzhao and Nie, Dujun and
Poggi, Matteo and Chen, Long},
booktitle={ICRA},
year={2025}
}

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 2.00 3.80 2.30
All / Est 2.00 3.80 2.30
Noc / All 1.83 3.52 2.11
Noc / Est 1.83 3.52 2.11
This table as LaTeX

Test Image 0

Error D1-bg D1-fg D1-all
All / All 2.12 0.57 1.90
All / Est 2.12 0.57 1.90
Noc / All 2.06 0.57 1.85
Noc / Est 2.06 0.57 1.85
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 1

Error D1-bg D1-fg D1-all
All / All 2.24 3.85 2.42
All / Est 2.24 3.85 2.42
Noc / All 2.18 3.85 2.37
Noc / Est 2.18 3.85 2.37
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 2

Error D1-bg D1-fg D1-all
All / All 3.03 6.96 3.22
All / Est 3.03 6.96 3.22
Noc / All 2.91 6.96 3.11
Noc / Est 2.91 6.96 3.11
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 3

Error D1-bg D1-fg D1-all
All / All 3.29 2.22 3.19
All / Est 3.29 2.22 3.19
Noc / All 3.23 2.22 3.14
Noc / Est 3.23 2.22 3.14
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 4

Error D1-bg D1-fg D1-all
All / All 1.25 2.83 1.51
All / Est 1.25 2.83 1.51
Noc / All 1.19 2.83 1.47
Noc / Est 1.19 2.83 1.47
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 5

Error D1-bg D1-fg D1-all
All / All 2.88 2.88 2.88
All / Est 2.88 2.88 2.88
Noc / All 2.64 2.88 2.67
Noc / Est 2.64 2.88 2.67
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 6

Error D1-bg D1-fg D1-all
All / All 4.45 1.01 4.09
All / Est 4.45 1.01 4.09
Noc / All 4.53 1.01 4.15
Noc / Est 4.53 1.01 4.15
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 7

Error D1-bg D1-fg D1-all
All / All 0.53 3.50 1.11
All / Est 0.53 3.50 1.11
Noc / All 0.54 3.50 1.13
Noc / Est 0.54 3.50 1.13
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 8

Error D1-bg D1-fg D1-all
All / All 0.54 2.19 0.84
All / Est 0.54 2.19 0.84
Noc / All 0.52 2.19 0.83
Noc / Est 0.52 2.19 0.83
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 9

Error D1-bg D1-fg D1-all
All / All 0.55 1.71 0.85
All / Est 0.55 1.71 0.85
Noc / All 0.55 1.80 0.86
Noc / Est 0.55 1.80 0.86
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 10

Error D1-bg D1-fg D1-all
All / All 1.63 2.16 1.75
All / Est 1.63 2.16 1.75
Noc / All 1.65 2.16 1.77
Noc / Est 1.65 2.16 1.77
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 11

Error D1-bg D1-fg D1-all
All / All 1.24 0.32 1.07
All / Est 1.24 0.32 1.07
Noc / All 1.24 0.32 1.08
Noc / Est 1.24 0.32 1.08
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 12

Error D1-bg D1-fg D1-all
All / All 0.65 0.60 0.65
All / Est 0.65 0.60 0.65
Noc / All 0.50 0.60 0.51
Noc / Est 0.50 0.60 0.51
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 13

Error D1-bg D1-fg D1-all
All / All 0.74 0.05 0.66
All / Est 0.74 0.05 0.66
Noc / All 0.60 0.05 0.54
Noc / Est 0.60 0.05 0.54
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 14

Error D1-bg D1-fg D1-all
All / All 1.42 0.11 1.40
All / Est 1.42 0.11 1.40
Noc / All 1.28 0.11 1.26
Noc / Est 1.28 0.11 1.26
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 15

Error D1-bg D1-fg D1-all
All / All 3.60 0.03 3.27
All / Est 3.60 0.03 3.27
Noc / All 3.67 0.03 3.33
Noc / Est 3.67 0.03 3.33
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 16

Error D1-bg D1-fg D1-all
All / All 4.14 0.09 3.54
All / Est 4.14 0.09 3.54
Noc / All 3.92 0.09 3.35
Noc / Est 3.92 0.09 3.35
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 17

Error D1-bg D1-fg D1-all
All / All 1.03 0.14 0.94
All / Est 1.03 0.14 0.94
Noc / All 1.04 0.14 0.94
Noc / Est 1.04 0.14 0.94
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 18

Error D1-bg D1-fg D1-all
All / All 5.64 0.88 3.38
All / Est 5.64 0.88 3.38
Noc / All 5.57 0.88 3.32
Noc / Est 5.57 0.88 3.32
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 19

Error D1-bg D1-fg D1-all
All / All 0.98 0.72 0.95
All / Est 0.98 0.72 0.95
Noc / All 0.97 0.72 0.94
Noc / Est 0.97 0.72 0.94
This table as LaTeX

Input Image

D1 Result

D1 Error




eXTReMe Tracker