computer-vision-7

3D Object Reconstruction

3260 papers • 126 benchmarks • 313 datasets

Image: Choy et al

(Image credit: Papersgraph)

Benchmarks

These leaderboards are used to track progress in 3d-object-reconstruction-7

Trend

Dataset

Best Model

Actions

Data3D−R2N2

BEHAVE

ShapeNet

Libraries

i

Use these libraries to find 3d-object-reconstruction-7 models and implementations

monniert/unicorn

2 papers 159

Datasets

ShapeNet

BEHAVE

X3D

Parcel3D

The RBO Dataset of Articulated Objects and Interactions

Subtasks

3D Object Reconstruction From A Single Image CAD Reconstruction

Most implemented papers

3D-R2N2: A Unified Approach for Single and Multi-view 3D Object Reconstruction

S. Savarese, JunYoung Gwak, C. Choy, Danfei Xu, Kevin Chen•Fri Apr 01 2016

The 3D-R2N2 reconstruction framework outperforms the state-of-the-art methods for single view reconstruction, and enables the 3D reconstruction of objects in situations when traditional SFM/SLAM methods fail (because of lack of texture and/or wide baseline).

1836

Content

RenderPeople

A collection of 131 CT datasets of pieces of modeling clay containing stones

Corn Seeds Dataset

0

Paper Graph

Pix2Vox: Context-Aware 3D Reconstruction From Single and Multi-View Images

Shengping Zhang, Shangchen Zhou, Haozhe Xie, H. Yao, Xiaoshuai Sun, Xiaojun Tong•Wed Jan 30 2019

Experimental results on the ShapeNet and Pix3D benchmarks indicate that the proposed Pix2Vox outperforms state-of-the-arts by a large margin, and the proposed method is 24 times faster than 3D-R2N2 in terms of backward inference time.

364 0

Paper Graph

Pixel2Mesh: Generating 3D Mesh Models from Single RGB Images

Yinda Zhang, Yu-Gang Jiang, Yanwei Fu, W. Liu, Nanyang Wang, Zhuwen Li•Wed Apr 04 2018

An end-to-end deep learning architecture that produces a 3D shape in triangular mesh from a single color image by progressively deforming an ellipsoid, leveraging perceptual features extracted from the input image.

1430 0

Paper Graph

A Point Set Generation Network for 3D Object Reconstruction from a Single Image

Hao Su, L. Guibas, Haoqiang Fan•Thu Dec 01 2016

This paper addresses the problem of 3D reconstruction from a single image, generating a straight-forward form of output unorthordox, and designs architecture, loss function and learning paradigm that are novel and effective, capable of predicting multiple plausible 3D point clouds from an input image.

2517 0

Paper Graph

3D Object Reconstruction from Hand-Object Interactions

Juergen Gall, Dimitrios Tzionas•Sun Dec 06 2015

This work shows that extracting 3d hand motion for in-hand scanning effectively facilitates the reconstruction of even featureless and highly symmetric objects and presents an approach that fuses the rich additional information of hands into a 3d reconstruction pipeline, significantly contributing to the state-of-the-art of in- hand scanning.

85 0

Paper Graph

Learning Efficient Point Cloud Generation for Dense 3D Object Reconstruction

S. Lucey, Chen-Hsuan Lin, Chen Kong•Tue Jun 20 2017

This paper uses 2D convolutional operations to predict the 3D structure from multiple viewpoints and jointly apply geometric reasoning with 2D projection optimization, and introduces the pseudo-renderer, a differentiable module to approximate the true rendering operation, to synthesize novel depth maps for optimization.

451 0

Paper Graph

Neural 3D Mesh Renderer

Y. Ushiku, T. Harada, Hiroharu Kato•Sun Nov 19 2017

This work proposes an approximate gradient for rasterization that enables the integration of rendering into neural networks and performs gradient-based 3D mesh editing operations, such as 2D-to-3D style transfer and 3D DeepDream, with 2D supervision for the first time.

1129 0

Paper Graph

Multi-View Silhouette and Depth Decomposition for High Resolution 3D Object Representation

Scott Fujimoto, D. Meger, Edward James Smith•Wed Jan 31 2018

This work introduces a novel method for the fast up-sampling of 3D objects in voxel space through networks that perform super-resolution on the six orthographic depth projections, and achieves state-of-the-art performance on 3D object reconstruction from RGB images on the ShapeNet dataset.

56 0

Paper Graph

Matryoshka Networks: Predicting 3D Geometry via Nested Shape Layers

Stephan R. Richter, S. Roth•Sat Apr 28 2018

Novel, efficient 2D encodings for 3D geometry enable reconstructing full 3D shapes from a single image at high resolution, and clearly outperform previous octree-based approaches despite having a much simpler architecture using standard network components.

143 0

Paper Graph

Pix2Vox++: Multi-scale Context-aware 3D Object Reconstruction from Single and Multiple Images

Shengping Zhang, Shangchen Zhou, Haozhe Xie, Wenxiu Sun, H. Yao•Sun Jun 21 2020

A novel framework for single-view and multi-view 3D object reconstruction, named Pix2Vox++ is proposed, which performs favorably against state-of-the-art methods in terms of both accuracy and efficiency.

211 0

Paper Graph

Adding a benchmark result helps the community track progress.