3D Reconstruction
559 papers with code • 8 benchmarks • 54 datasets
3D Reconstruction is the task of creating a 3D model or representation of an object or scene from 2D images or other data sources. The goal of 3D reconstruction is to create a virtual representation of an object or scene that can be used for a variety of purposes, such as visualization, animation, simulation, and analysis. It can be used in fields such as computer vision, robotics, and virtual reality.
Image: Gwak et al
Benchmarks
These leaderboards are used to track progress in 3D Reconstruction
Libraries
Use these libraries to find 3D Reconstruction models and implementations
ShapeNet
ScanNet
DTU
300W
ShapeNetCore
MegaDepth
ABC Dataset
ETH3D
BlendedMVS
Scan2CAD
Subtasks
Most implemented papers
Instant Neural Graphics Primitives with a Multiresolution Hash Encoding
Neural graphics primitives, parameterized by fully connected neural networks, can be costly to train and evaluate.
3D-R2N2: A Unified Approach for Single and Multi-view 3D Object Reconstruction
Inspired by the recent success of methods that employ shape priors to achieve robust 3D reconstructions, we propose a novel recurrent neural network architecture that we call the 3D Recurrent Reconstruction Neural Network (3D-R2N2).
Kimera: an Open-Source Library for Real-Time Metric-Semantic Localization and Mapping
MIT-SPARK/Kimera
•
•
We provide an open-source C++ library for real-time metric-semantic visual-inertial Simultaneous Localization And Mapping (SLAM).
The Double Sphere Camera Model
We evaluate the model using a calibration dataset with several different lenses and compare the models using the metrics that are relevant for Visual Odometry, i. e., reprojection error, as well as computation time for projection and unprojection functions and their Jacobians.
Occupancy Networks: Learning 3D Reconstruction in Function Space
LMescheder/Occupancy-Networks
•
•
CVPR 2019
With the advent of deep neural networks, learning-based approaches for 3D reconstruction have gained popularity.
PCRNet: Point Cloud Registration Network using PointNet Encoding
PointNet has recently emerged as a popular representation for unstructured point cloud data, allowing application of deep learning to tasks such as object detection, segmentation and shape completion.
Convolutional Occupancy Networks
autonomousvision/convolutional_occupancy_networks
•
•
ECCV 2020
Recently, implicit neural representations have gained popularity for learning-based 3D reconstruction.
EPP-MVSNet: Epipolar-Assembling Based Depth Prediction for Multi-View Stereo
mindspore-ai/models
•
•
ICCV 2021
As a result, we achieve promising results on all datasets and the highest F-Score on the online TNT intermediate benchmark.
Pix2Vox: Context-aware 3D Reconstruction from Single and Multi-view Images
Then, a context-aware fusion module is introduced to adaptively select high-quality reconstructions for each part (e. g., table legs) from different coarse 3D volumes to obtain a fused 3D volume.
Incremental Visual-Inertial 3D Mesh Generation with Structural Regularities
MIT-SPARK/Kimera
•
•
We propose instead to tightly couple mesh regularization and state estimation by detecting and enforcing structural regularities in a novel factor-graph formulation.
