List your company

Depth Completion on Matterport3D (test)

0.5223RMSEGLPanoDepth(Complete)
0.4905520.7048510.919151.133449May 18, 2020Aug 30, 2020Dec 13, 2020Mar 28, 2021Jul 11, 2021Oct 24, 2021Feb 6, 2022

Evaluation results

MethodLinks
GLPanoDepth(Complete)Method architecture=Full architecture2022.020.52230.2998——86.4195.6198.08—0.0786
GLPanoDepth(ViT+CNN)Method architecture=ViT and CNN hybrid2022.020.54660.3232——84.7695.1297.93—0.0827
GLPanoDepth(Concat)Method architecture=Concatenation fusion2022.020.57470.3407——82.8894.6697.79—0.0862
SliceNetMethod architecture=standard2022.020.61330.3296——87.1694.8397.16—0.1045
BiFuseMethod architecture=standard2022.020.62590.347——84.5293.1996.32—0.1134
FCRNMethod architecture=standard2022.020.67040.4008——77.0391.7496.17—0.1244
OmniDepthMethod architecture=standard2022.020.76430.4838——68.387.9494.29—0.145
DM-LRNdecoder modulation branch=included2020.050.9610.28572.681.38993.394.90.844—
LRNdecoder modulation branch=none, mask on input=none2020.051.0280.29971.980.58993.2950.843—
Li et al.training=official implementation2020.051.0540.39750.863.177.587.4920.7—
LRN + maskdecoder modulation branch=none, mask on input=included2020.051.0540.29873.781.588.993.3950.844—
Huang et al.source=Reported in [16], model=official pretrained2020.051.0920.34266.1758591.193.60.799—
Li et al.training pipeline=proposed pipeline2020.051.1340.42664.972.983.489.992.80.774—
Gansbeke et al.training=official implementation2020.051.1610.39554.265.779.988.792.70.7—
Gansbeke et al.training pipeline=proposed pipeline2020.051.2640.48467.574.182.688.8920.78—
Zhang et al.source=Reported in [16]2020.051.3160.46165.770.878.185.188.80.762—