Home Console Docs News Papers Tutorials Datasets Wiki SOTA LLM Models GPU Leaderboard Events

English

Depth Estimation On Stanford2D3D Panoramic

Metrics

RMSE

absolute relative error

Results

Performance results of various models on this benchmark

			Paper Title	Repository
OmniDepth	0.6152	0.1996	OmniDepth: Dense Depth Estimation for Indoors Spherical Panoramas
SphereDepth	0.4512	0.1158	SphereDepth: Panorama Depth Estimation from Spherical Domain	-
Jin et al.	0.421	-	Geometric Structure Based and Regularized Depth Estimation From 360 Indoor Imagery	-
BiFuse with fusion	0.4142	0.1209	BiFuse: Monocular 360 Depth Estimation via Bi-Projection Fusion	-
HoHoNet (ResNet-101)	0.3834	0.1014	HoHoNet: 360 Indoor Holistic Understanding with Latent Horizontal Features
PanoDepth	0.3747	0.0972	PanoDepth: A Two-Stage Approach for Monocular Omnidirectional Depth Estimation	-
BiFuse++	0.372	0.1117	BiFuse++: Self-supervised and Efficient Bi-projection Fusion for 360 Depth Estimation
UniFuse with fusion	0.3691	0.1114	UniFuse: Unidirectional Fusion for 360$^{\circ}$ Panorama Depth Estimation
DisConv	0.369	0.176	Distortion-Aware Convolutional Filters for Dense Prediction in Panoramic Images	-
SliceNet	0.3684	0.0744	SliceNet: Deep Dense Depth Estimation From a Single Indoor Panorama Using a Slice-Based Representation	-
Neural Contourlet Network	0.3528	0.0558	Neural Contourlet Network for Monocular 360 Depth Estimation
GLPanoDepth	0.3493	-	GLPanoDepth: Global-to-Local Panoramic Depth Estimation
OmniFusion (2-iter)	0.3474	0.095	OmniFusion: 360 Monocular Depth Estimation via Geometry-Aware Fusion
ACDNet	0.341	0.0984	ACDNet: Adaptively Combined Dilated Convolution for Monocular Panorama Depth Estimation
PanoFormer	0.3083	0.0405	PanoFormer: Panorama Transformer for Indoor 360 Depth Estimation
NLFB	0.2776	0.0649	Improving 360 Monocular Depth Estimation via Non-local Dense Prediction Transformer and Joint Supervised and Self-supervised Learning
FreDSNet	0.2727	0.0952	FreDSNet: Joint Monocular Depth and Semantic Segmentation with Fast Fourier Convolutions
HiMODE	0.2619	0.0532	HiMODE: A Hybrid Monocular Omnidirectional Depth Estimation Model	-

0 of 18 row(s) selected.

Depth Estimation On Stanford2D3D Panoramic | SOTA | HyperAI