Method

2D-3D FuseNet [UberATG-FuseNet]


Submitted on 7 Aug. 2019 17:18 by
Yun Chen (Uber ATG)

Running time:0.09 s
Environment:GPU @ 2.5 Ghz (Python)

Method Description:
In this paper, we tackle the problem of depth
completion from RGBD data. Towards this goal, we
design a simple yet effective neural network
block that learns to extract joint 2D and 3D
features. Specifically, the block consists of two
domain-specific sub-networks that apply 2D
convolution on image pixels and continuous
convolution on 3D points, with their output
features fused in image space. We build the depth
completion network simply by stacking the
proposed block, which has the advantage of
learning hierarchical representations that are
fully fused between 2D and 3D spaces at multiple
levels. We demonstrate the effectiveness of our
approach on the challenging KITTI depth
completion benchmark and show that our approach
outperforms the state-of-the-art.
Parameters:
See the paper
Latex Bibtex:
@inproceedings{learning2019yun,
author = {Yun Chen and
Bin Yang and
Ming Liang and
Raquel Urtasun},
title = {Learning Joint 2D-3D Representations
for Depth Completion},
booktitle = {ICCV},
year = {2019}}

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 Sparsity Invariant CNNs (THREEDV 2017), 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 Sparsity Invariant CNNs (THREEDV 2017), 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

iRMSE iMAE RMSE MAE
Error 2.34 1.14 752.88 221.19
This table as LaTeX

Test Image 0

iRMSE iMAE RMSE MAE
Error 2.84 0.85 790.21 174.31
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 1

iRMSE iMAE RMSE MAE
Error 2.95 1.19 744.37 104.69
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 2

iRMSE iMAE RMSE MAE
Error 2.44 1.69 1180.63 444.18
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 3

iRMSE iMAE RMSE MAE
Error 3.67 1.99 625.98 260.78
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 4

iRMSE iMAE RMSE MAE
Error 3.10 1.62 577.25 223.41
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 5

iRMSE iMAE RMSE MAE
Error 3.52 1.19 765.10 178.66
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 6

iRMSE iMAE RMSE MAE
Error 3.43 1.23 436.64 166.80
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 7

iRMSE iMAE RMSE MAE
Error 4.67 1.74 615.33 182.58
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 8

iRMSE iMAE RMSE MAE
Error 1.88 0.74 726.99 163.93
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 9

iRMSE iMAE RMSE MAE
Error 2.28 1.26 779.20 231.14
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 10

iRMSE iMAE RMSE MAE
Error 2.10 1.46 757.58 407.05
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 11

iRMSE iMAE RMSE MAE
Error 2.48 1.19 1176.05 396.35
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 12

iRMSE iMAE RMSE MAE
Error 4.41 1.98 961.33 256.98
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 13

iRMSE iMAE RMSE MAE
Error 1.42 0.96 589.63 198.90
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 14

iRMSE iMAE RMSE MAE
Error 1.76 1.06 558.69 158.06
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 15

iRMSE iMAE RMSE MAE
Error 3.85 1.76 470.12 179.64
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 16

iRMSE iMAE RMSE MAE
Error 1.38 0.79 548.33 178.08
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 17

iRMSE iMAE RMSE MAE
Error 1.48 0.81 588.71 176.82
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 18

iRMSE iMAE RMSE MAE
Error 1.73 0.89 661.42 247.80
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 19

iRMSE iMAE RMSE MAE
Error 1.45 1.06 718.93 237.06
This table as LaTeX

Input Image

D1 Result

D1 Error




eXTReMe Tracker