Method

Displets: Resolving Stereo Ambiguities using Object Knowledge [Displets]
http://www.cvlibs.net/projects/displets/

Submitted on 30 Mar. 2015 18:29 by
Fatma Guney (MPI)

Running time:265 s
Environment:>8 cores @ 3.0 Ghz (Matlab + C/C++)

Method Description:
Stereo techniques have witnessed tremendous progress over the last decades, yet some aspects of the problem still remain challenging today. Striking examples are reflecting and textureless surfaces which cannot easily be recovered using traditional local regularizers. In this paper, we therefore propose to regularize over larger distances using object-category specific disparity proposals (displets) which we sample using inverse graphics techniques based on a sparse disparity estimate and a semantic segmentation of the image. The proposed displets encode the fact that objects of certain categories are not arbitrarily shaped but typically exhibit regular structures. We integrate them as non-local regularizer for the challenging object class 'car' into a superpixel based CRF framework.
Parameters:
Number of Particles: 30
Number of Superpixels: 1000
Number of Iterations: 40
TRWS - Number of Iterations: 200
Particle Standard Deviations: [0.5 0.5 5]

Please check Table 1 of the supplementary material for a complete list of parameters.
Latex Bibtex:
@INPROCEEDINGS{Guney2015CVPR,
author = {Fatma Guney and Andreas Geiger},
title = {Displets: Resolving Stereo Ambiguities using Object Knowledge},
booktitle = {Conference on Computer Vision and Pattern Recognition (CVPR)},
year = {2015}
}

Detailed Results

This page provides detailed results for the method(s) selected. For each of the first 20 test images, the number of erroneous pixels at all thresholds is depicted in the table. Underneath, the left input image, the disparity / end-point error map and the estimated (and interpolated) disparity / optical flow map are shown. The error map scales linearly between 0 (black) and >=5 (white) pixels error. Red denotes all occluded pixels, falling outside the image boundaries. The false color map is scaled to the largest ground truth disparity / flow value.

Test Set Average

Error Out-Noc Out-All Avg-Noc Avg-All
2 pixels 3.90 % 4.92 % 0.7 px 0.9 px
3 pixels 2.47 % 3.27 % 0.7 px 0.9 px
4 pixels 1.94 % 2.62 % 0.7 px 0.9 px
5 pixels 1.67 % 2.25 % 0.7 px 0.9 px
This table as LaTeX

Reflective Regions

Error Out-Noc Out-All Avg-Noc Avg-All
2 pixels 14.87 % 16.76 % 1.9 px 2.3 px
3 pixels 8.40 % 9.89 % 1.9 px 2.3 px
4 pixels 6.11 % 7.23 % 1.9 px 2.3 px
5 pixels 5.07 % 5.94 % 1.9 px 2.3 px
This table as LaTeX

Test Image 0

Error Out-Noc Out-All Avg-Noc Avg-All
2 pixels 3.64 % 3.75 % 0.7 px 0.7 px
3 pixels 2.33 % 2.48 % 0.7 px 0.7 px
4 pixels 1.97 % 2.13 % 0.7 px 0.7 px
5 pixels 1.71 % 1.88 % 0.7 px 0.7 px
This table as LaTeX





Test Image 1

Error Out-Noc Out-All Avg-Noc Avg-All
2 pixels 3.90 % 5.23 % 0.7 px 0.8 px
3 pixels 2.17 % 3.15 % 0.7 px 0.8 px
4 pixels 1.84 % 2.75 % 0.7 px 0.8 px
5 pixels 1.65 % 2.53 % 0.7 px 0.8 px
This table as LaTeX





Test Image 2

Error Out-Noc Out-All Avg-Noc Avg-All
2 pixels 2.90 % 3.80 % 0.5 px 0.6 px
3 pixels 1.56 % 2.38 % 0.5 px 0.6 px
4 pixels 1.17 % 1.86 % 0.5 px 0.6 px
5 pixels 0.96 % 1.52 % 0.5 px 0.6 px
This table as LaTeX





Test Image 3

Error Out-Noc Out-All Avg-Noc Avg-All
2 pixels 3.24 % 3.78 % 0.6 px 0.6 px
3 pixels 1.16 % 1.21 % 0.6 px 0.6 px
4 pixels 0.81 % 0.88 % 0.6 px 0.6 px
5 pixels 0.67 % 0.74 % 0.6 px 0.6 px
This table as LaTeX





Test Image 4

Error Out-Noc Out-All Avg-Noc Avg-All
2 pixels 2.72 % 3.87 % 0.5 px 0.6 px
3 pixels 1.74 % 1.89 % 0.5 px 0.6 px
4 pixels 1.45 % 1.55 % 0.5 px 0.6 px
5 pixels 1.26 % 1.37 % 0.5 px 0.6 px
This table as LaTeX





Test Image 5

Error Out-Noc Out-All Avg-Noc Avg-All
2 pixels 2.08 % 2.29 % 0.5 px 0.5 px
3 pixels 1.42 % 1.57 % 0.5 px 0.5 px
4 pixels 1.26 % 1.37 % 0.5 px 0.5 px
5 pixels 1.16 % 1.22 % 0.5 px 0.5 px
This table as LaTeX





Test Image 6

Error Out-Noc Out-All Avg-Noc Avg-All
2 pixels 4.96 % 7.44 % 0.8 px 0.9 px
3 pixels 2.43 % 4.02 % 0.8 px 0.9 px
4 pixels 1.84 % 2.32 % 0.8 px 0.9 px
5 pixels 1.57 % 1.95 % 0.8 px 0.9 px
This table as LaTeX





Test Image 7

Error Out-Noc Out-All Avg-Noc Avg-All
2 pixels 5.10 % 5.57 % 1.0 px 1.1 px
3 pixels 3.69 % 4.00 % 1.0 px 1.1 px
4 pixels 3.12 % 3.40 % 1.0 px 1.1 px
5 pixels 2.49 % 2.75 % 1.0 px 1.1 px
This table as LaTeX





Test Image 8

Error Out-Noc Out-All Avg-Noc Avg-All
2 pixels 1.86 % 2.30 % 0.5 px 0.5 px
3 pixels 1.08 % 1.54 % 0.5 px 0.5 px
4 pixels 0.86 % 1.32 % 0.5 px 0.5 px
5 pixels 0.77 % 1.23 % 0.5 px 0.5 px
This table as LaTeX





Test Image 9

Error Out-Noc Out-All Avg-Noc Avg-All
2 pixels 6.01 % 8.31 % 1.1 px 1.5 px
3 pixels 3.24 % 5.58 % 1.1 px 1.5 px
4 pixels 2.23 % 4.39 % 1.1 px 1.5 px
5 pixels 1.91 % 3.61 % 1.1 px 1.5 px
This table as LaTeX





Test Image 10

Error Out-Noc Out-All Avg-Noc Avg-All
2 pixels 2.37 % 3.90 % 0.5 px 0.6 px
3 pixels 1.54 % 2.91 % 0.5 px 0.6 px
4 pixels 1.20 % 2.27 % 0.5 px 0.6 px
5 pixels 1.12 % 1.99 % 0.5 px 0.6 px
This table as LaTeX





Test Image 11

Error Out-Noc Out-All Avg-Noc Avg-All
2 pixels 3.16 % 3.48 % 1.0 px 1.0 px
3 pixels 1.91 % 2.26 % 1.0 px 1.0 px
4 pixels 1.64 % 2.01 % 1.0 px 1.0 px
5 pixels 1.44 % 1.81 % 1.0 px 1.0 px
This table as LaTeX





Test Image 12

Error Out-Noc Out-All Avg-Noc Avg-All
2 pixels 4.98 % 6.97 % 1.2 px 1.3 px
3 pixels 3.89 % 5.47 % 1.2 px 1.3 px
4 pixels 3.58 % 4.78 % 1.2 px 1.3 px
5 pixels 3.47 % 4.42 % 1.2 px 1.3 px
This table as LaTeX





Test Image 13

Error Out-Noc Out-All Avg-Noc Avg-All
2 pixels 3.66 % 4.03 % 0.9 px 0.9 px
3 pixels 2.79 % 2.81 % 0.9 px 0.9 px
4 pixels 2.50 % 2.47 % 0.9 px 0.9 px
5 pixels 2.28 % 2.26 % 0.9 px 0.9 px
This table as LaTeX





Test Image 14

Error Out-Noc Out-All Avg-Noc Avg-All
2 pixels 0.96 % 1.46 % 0.4 px 0.5 px
3 pixels 0.63 % 0.77 % 0.4 px 0.5 px
4 pixels 0.57 % 0.69 % 0.4 px 0.5 px
5 pixels 0.51 % 0.63 % 0.4 px 0.5 px
This table as LaTeX





Test Image 15

Error Out-Noc Out-All Avg-Noc Avg-All
2 pixels 4.46 % 4.54 % 0.7 px 0.7 px
3 pixels 2.60 % 2.53 % 0.7 px 0.7 px
4 pixels 1.66 % 1.62 % 0.7 px 0.7 px
5 pixels 1.17 % 1.14 % 0.7 px 0.7 px
This table as LaTeX





Test Image 16

Error Out-Noc Out-All Avg-Noc Avg-All
2 pixels 4.19 % 4.13 % 0.7 px 0.7 px
3 pixels 1.34 % 1.32 % 0.7 px 0.7 px
4 pixels 0.70 % 0.69 % 0.7 px 0.7 px
5 pixels 0.48 % 0.47 % 0.7 px 0.7 px
This table as LaTeX





Test Image 17

Error Out-Noc Out-All Avg-Noc Avg-All
2 pixels 5.77 % 6.96 % 0.8 px 0.9 px
3 pixels 4.27 % 5.18 % 0.8 px 0.9 px
4 pixels 3.61 % 4.15 % 0.8 px 0.9 px
5 pixels 2.82 % 3.08 % 0.8 px 0.9 px
This table as LaTeX





Test Image 18

Error Out-Noc Out-All Avg-Noc Avg-All
2 pixels 1.84 % 2.19 % 0.4 px 0.4 px
3 pixels 0.79 % 0.91 % 0.4 px 0.4 px
4 pixels 0.41 % 0.50 % 0.4 px 0.4 px
5 pixels 0.33 % 0.35 % 0.4 px 0.4 px
This table as LaTeX





Test Image 19

Error Out-Noc Out-All Avg-Noc Avg-All
2 pixels 1.52 % 1.50 % 0.5 px 0.5 px
3 pixels 0.87 % 0.84 % 0.5 px 0.5 px
4 pixels 0.71 % 0.69 % 0.5 px 0.5 px
5 pixels 0.65 % 0.63 % 0.5 px 0.5 px
This table as LaTeX







eXTReMe Tracker