3D Reconstruction On Dtu

Metrics

Acc

Comp

Overall

Results

Performance results of various models on this benchmark

				Paper Title	Repository
PatchmatchNet	0.427	0.277	0.352	PatchmatchNet: Learned Multi-View Patchmatch Stereo
EPP-MVSNet	0.413	0.296	0.355	EPP-MVSNet: Epipolar-Assembling Based Depth Prediction for Multi-View Stereo	-
COLMAP	0.400	0.664	0.532	Structure-From-Motion Revisited	-
3D-R2N2	0.397	0.884	0.630	3D-R2N2: A Unified Approach for Single and Multi-view 3D Object Reconstruction
MVSNet	0.396	0.527	0.462	MVSNet: Depth Inference for Unstructured Multi-view Stereo
MSCVP-MVSNet	0.379	0.278	0.328	Cost Volume Pyramid Network with Multi-strategies Range Searching for Multi-view Stereo
AA-RMVSNet	0.376	0.339	0.357	AA-RMVSNet: Adaptive Aggregation Recurrent Multi-view Stereo Network
Vis-MVSNet	0.369	0.361	0.365	Visibility-aware Multi-view Stereo Network
UniMVSNet	0.352	0.278	0.315	Rethinking Depth Estimation for Multi-View Stereo: A Unified Representation
CDS-MVSNet	0.351	0.278	0.315	Curvature-guided dynamic scale networks for Multi-view Stereo
GoMVS	0.347	0.227	0.287	GoMVS: Geometrically Consistent Cost Aggregation for Multi-View Stereo
UCSNet	0.338	0.349	0.344	Deep Stereo using Adaptive Thin Volume Representation with Uncertainty Awareness
IB-MVS	0.334	0.309	0.321	IB-MVS: An Iterative Algorithm for Deep Multi-View Stereo based on Binary Decisions	-
GeoMVSNet	0.331	0.259	0.295	GeoMVSNet: Learning Multi-View Stereo With Geometry Perception	-
GC-MVSNet	0.330	0.260	0.295	GC-MVSNet: Multi-View, Multi-Scale, Geometrically-Consistent Multi-View Stereo
ET-MVSNet	0.329	0.253	0.291	When Epipolar Constraint Meets Non-local Operators in Multi-View Stereo
MVSFormer	0.327	0.251	0.289	MVSFormer: Multi-View Stereo by Learning Robust Image Features and Temperature-based Depth
RA-MVSNet	0.326	0.268	0.297	Multi-View Stereo Representation Revisit: Region-Aware MVSNet	-
Cas-MVSNet	0.325	0.385	0.355	Cascade Cost Volume for High-Resolution Multi-View Stereo and Stereo Matching
TransMVSNet	0.321	0.289	0.305	TransMVSNet: Global Context-aware Multi-view Stereo Network with Transformers

0 of 24 row(s) selected.

Command Palette

3D Reconstruction On Dtu

Metrics

Results