Method

Pretrained-based Scale Recovery for Depth Completion [PS-Net]


Submitted on 10 Mar. 2025 17:11 by
Li Bihan (Zhe Jiang University)

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

Method Description:
The method completes sparse LiDAR depth using
geometric cues from RGB images to obtain a dense
and accurate real depth map. Computation is
performed in the scale space through scale
decomposition. In the scale recovery module, the
method first utilizes cross-attention for pre-
filling the sparse scale based on the spatial
similarity between local points in the image. Then,
the image features extracted by the pretrained
encoder are fused with the scale features extracted
from the scale map for decoding, improving
robustness across different scenes. Finally,
through adaptive adjustment of neighborhood
selection and affinity computation in the spatial
propagation network, the scale map undergoes
iterative refinement, yielding more accurate scale
results.
Parameters:
27.78M
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 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 2.01 0.89 698.00 202.04
This table as LaTeX

Test Image 0

iRMSE iMAE RMSE MAE
Error 2.60 0.67 744.50 156.07
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 1

iRMSE iMAE RMSE MAE
Error 2.26 0.73 584.78 72.40
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 2

iRMSE iMAE RMSE MAE
Error 2.08 1.45 1130.69 424.45
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 3

iRMSE iMAE RMSE MAE
Error 3.46 1.57 616.72 235.17
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 4

iRMSE iMAE RMSE MAE
Error 2.68 1.40 510.17 210.34
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 5

iRMSE iMAE RMSE MAE
Error 2.80 0.87 803.32 162.60
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 6

iRMSE iMAE RMSE MAE
Error 3.61 0.89 738.18 159.53
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 7

iRMSE iMAE RMSE MAE
Error 3.83 1.18 633.42 162.81
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 8

iRMSE iMAE RMSE MAE
Error 1.84 0.62 668.65 148.50
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 9

iRMSE iMAE RMSE MAE
Error 1.87 1.05 641.75 200.02
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 10

iRMSE iMAE RMSE MAE
Error 1.85 1.31 763.86 399.48
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 11

iRMSE iMAE RMSE MAE
Error 2.23 1.03 1128.48 381.19
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 12

iRMSE iMAE RMSE MAE
Error 3.35 1.39 963.95 238.10
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 13

iRMSE iMAE RMSE MAE
Error 1.23 0.77 650.77 192.66
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 14

iRMSE iMAE RMSE MAE
Error 1.37 0.77 536.67 138.52
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 15

iRMSE iMAE RMSE MAE
Error 3.82 1.27 489.85 158.50
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 16

iRMSE iMAE RMSE MAE
Error 1.17 0.64 514.94 161.78
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 17

iRMSE iMAE RMSE MAE
Error 1.24 0.63 548.44 161.37
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 18

iRMSE iMAE RMSE MAE
Error 1.54 0.80 597.75 233.20
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 19

iRMSE iMAE RMSE MAE
Error 1.08 0.78 655.04 212.87
This table as LaTeX

Input Image

D1 Result

D1 Error




eXTReMe Tracker