Method

Bi3D: Stereo Depth Estimation via Binary Classifications [Bi3D]
https://github.com/NVlabs/Bi3D

Submitted on 18 May. 2020 20:38 by
abhishek badki (UCSB)

Running time:0.48 s
Environment:GPU @ 1.5 Ghz (Python)

Method Description:
Bi3D estimates depth via a series of binary
classifications. Rather than testing if objects are at a
particular depth D, as existing stereo methods do, it
classifies them as being closer or farther than D. This
property offers a powerful mechanism to balance accuracy
and latency. Given a strict time budget, Bi3D can detect
objects closer than a given distance in as little as a
few milliseconds, or estimate depth with arbitrarily
coarse quantization, with complexity linear with the
number of quantization levels. Bi3D can also use the
allotted quantization levels to get continuous depth, but
in a specific depth range.
Parameters:
SPP feature extraction for feature extraction,
standard 2D auto-encoder with skip connections for
multiple binary classifications, simple disparity
refinement network
Latex Bibtex:
@InProceedings{badki2020Bi3D,
author = {Badki, Abhishek and Troccoli, Alejandro and Kim, Kihwan and Kautz, Jan and Sen, Pradeep and Gallo, Orazio},
title = {Bi3D: Stereo Depth Estimation via Binary Classifications},
booktitle = {The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
year = {2020}
}

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 1.95 3.48 2.21
All / Est 1.95 3.48 2.21
Noc / All 1.79 3.11 2.01
Noc / Est 1.79 3.11 2.01
This table as LaTeX

Test Image 0

Error D1-bg D1-fg D1-all
All / All 2.09 0.92 1.93
All / Est 2.09 0.92 1.93
Noc / All 2.04 0.92 1.88
Noc / Est 2.04 0.92 1.88
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 1

Error D1-bg D1-fg D1-all
All / All 2.06 2.43 2.10
All / Est 2.06 2.43 2.10
Noc / All 1.96 2.43 2.01
Noc / Est 1.96 2.43 2.01
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 2

Error D1-bg D1-fg D1-all
All / All 3.05 5.86 3.19
All / Est 3.05 5.86 3.19
Noc / All 2.84 5.86 2.99
Noc / Est 2.84 5.86 2.99
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 3

Error D1-bg D1-fg D1-all
All / All 2.67 1.31 2.55
All / Est 2.67 1.31 2.55
Noc / All 2.61 1.31 2.49
Noc / Est 2.61 1.31 2.49
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 4

Error D1-bg D1-fg D1-all
All / All 2.07 0.52 1.81
All / Est 2.07 0.52 1.81
Noc / All 1.86 0.52 1.64
Noc / Est 1.86 0.52 1.64
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 5

Error D1-bg D1-fg D1-all
All / All 2.69 2.15 2.64
All / Est 2.69 2.15 2.64
Noc / All 2.32 2.15 2.31
Noc / Est 2.32 2.15 2.31
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 6

Error D1-bg D1-fg D1-all
All / All 5.29 0.50 4.78
All / Est 5.29 0.50 4.78
Noc / All 5.40 0.50 4.88
Noc / Est 5.40 0.50 4.88
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 7

Error D1-bg D1-fg D1-all
All / All 0.41 3.13 0.94
All / Est 0.41 3.13 0.94
Noc / All 0.41 3.13 0.95
Noc / Est 0.41 3.13 0.95
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 8

Error D1-bg D1-fg D1-all
All / All 0.34 1.92 0.63
All / Est 0.34 1.92 0.63
Noc / All 0.34 1.92 0.63
Noc / Est 0.34 1.92 0.63
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 9

Error D1-bg D1-fg D1-all
All / All 0.39 1.29 0.62
All / Est 0.39 1.29 0.62
Noc / All 0.39 1.35 0.63
Noc / Est 0.39 1.35 0.63
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 10

Error D1-bg D1-fg D1-all
All / All 1.63 2.52 1.83
All / Est 1.63 2.52 1.83
Noc / All 1.65 2.52 1.85
Noc / Est 1.65 2.52 1.85
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 11

Error D1-bg D1-fg D1-all
All / All 0.96 0.26 0.84
All / Est 0.96 0.26 0.84
Noc / All 0.97 0.26 0.84
Noc / Est 0.97 0.26 0.84
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 12

Error D1-bg D1-fg D1-all
All / All 0.68 0.44 0.66
All / Est 0.68 0.44 0.66
Noc / All 0.54 0.44 0.53
Noc / Est 0.54 0.44 0.53
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 13

Error D1-bg D1-fg D1-all
All / All 0.84 0.21 0.76
All / Est 0.84 0.21 0.76
Noc / All 0.81 0.21 0.74
Noc / Est 0.81 0.21 0.74
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 14

Error D1-bg D1-fg D1-all
All / All 1.63 0.00 1.60
All / Est 1.63 0.00 1.60
Noc / All 1.43 0.00 1.41
Noc / Est 1.43 0.00 1.41
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 15

Error D1-bg D1-fg D1-all
All / All 3.47 0.11 3.16
All / Est 3.47 0.11 3.16
Noc / All 3.54 0.11 3.22
Noc / Est 3.54 0.11 3.22
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 16

Error D1-bg D1-fg D1-all
All / All 3.57 0.03 3.05
All / Est 3.57 0.03 3.05
Noc / All 3.37 0.03 2.88
Noc / Est 3.37 0.03 2.88
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 17

Error D1-bg D1-fg D1-all
All / All 0.88 0.23 0.82
All / Est 0.88 0.23 0.82
Noc / All 0.89 0.23 0.82
Noc / Est 0.89 0.23 0.82
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 18

Error D1-bg D1-fg D1-all
All / All 4.92 0.88 3.00
All / Est 4.92 0.88 3.00
Noc / All 4.80 0.88 2.92
Noc / Est 4.80 0.88 2.92
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 19

Error D1-bg D1-fg D1-all
All / All 0.78 0.10 0.70
All / Est 0.78 0.10 0.70
Noc / All 0.78 0.10 0.70
Noc / Est 0.78 0.10 0.70
This table as LaTeX

Input Image

D1 Result

D1 Error




eXTReMe Tracker