Filters








328,325 Hits in 5.8 sec

3D Reconstruction of Simple Objects from A Single View Silhouette Image [article]

Xinhan Di, Pengqian Yu
2017 arXiv   pre-print
Evaluation is performed using Shapenet for the single-view reconstruction and results are presented in comparison with a single network, to highlight the improvements obtained with the proposed stacked  ...  Furthermore, 3D re- construction in forms of IoU is compared with the state of art 3D reconstruction from a single-view RGB image, and the proposed model achieves higher IoU than the state of art of reconstruction  ...  The third column represents the rebuilt 3D shape reconstructed from the proposed stacked networks. The fourth column represents the ground truth shape.  ... 
arXiv:1701.04752v1 fatcat:d7czonac5za23gpgvngj6sjcmi

A Survey on Deep Geometry Learning: From a Representation Perspective [article]

Yun-Peng Xiao, Yu-Kun Lai, Fang-Lue Zhang, Chunpeng Li, Lin Gao
2020 arXiv   pre-print
Unlike 2D images, which can be uniformly represented by regular grids of pixels, 3D shapes have various representations, such as depth and multi-view images, voxel-based representation, point-based representation  ...  Many advanced techniques for 3D shapes have been proposed for different applications.  ...  A primitive representation represents the 3D shape with primitives such as oriented 3D boxes.  ... 
arXiv:2002.07995v2 fatcat:pustwlu5freypnccfrculkqvei

A survey on deep geometry learning: From a representation perspective

Yun-Peng Xiao, Yu-Kun Lai, Fang-Lue Zhang, Chunpeng Li, Lin Gao
2020 Computational Visual Media  
Unlike 2D images, which can be uniformly represented by a regular grid of pixels, 3D shapes have various representations, such as depth images, multi-view images, voxels, point clouds, meshes, implicit  ...  Many advanced techniques for 3D shapes have been proposed for different applications.  ...  They provide a description with infinite resolution for 3D shapes with reasonable memory consumption, and are capable of representing shapes with changing topology.  ... 
doi:10.1007/s41095-020-0174-8 fatcat:kpoynaixq5esbek63bovybisfa

A compact and scalable representation of network traffic dynamics using shapes and its applications

Panchamy Krishnakumari, Oded Cats, Hans van Lint
2020 Transportation Research Part C: Emerging Technologies  
This is done by extracting pockets of congestion that encompass connected 3D subnetworks as 3D shapes.  ...  Moreover, these partitioning techniques was used to synthesize the network of Amsterdam over 35 days into 4 so-called consensual patterns with each 9 homogeneous 3D subnetworks and show that with these  ...  A threshold of 0 implies that only links with the same Fig. 2 . 3D network with unidirectional virtual links connecting the 2D networks and representing time t where different colors display the edge  ... 
doi:10.1016/j.trc.2020.102850 fatcat:jvbmwyruzbdixgid66joxm7qsu

DeepSDF: Learning Continuous Signed Distance Functions for Shape Representation

Jeong Joon Park, Peter Florence, Julian Straub, Richard Newcombe, Steven Lovegrove
2019 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)  
Figure 1 : DeepSDF represents signed distance functions (SDFs) of shapes via latent code-conditioned feed-forward decoder networks.  ...  Furthermore, we show stateof-the-art performance for learned 3D shape representation and completion while reducing the model size by an order of magnitude compared with previous work.  ...  Our tasks evaluated are: (K) representing known shapes, (U) representing unknown shapes, and (C) shape completion. mesh with proper orientation, we set up equally spaced virtual cameras around the object  ... 
doi:10.1109/cvpr.2019.00025 dblp:conf/cvpr/ParkFSNL19 fatcat:6wnfa36lqvbxdbnrc3het6mfbe

Part2Word: Learning Joint Embedding of Point Clouds and Text by Matching Parts to Words [article]

Chuan Tang, Xi Yang, Bojian Wu, Zhizhong Han, Yi Chang
2021 arXiv   pre-print
In the optimized space, we represent a part by aggregating features of all points within the part, while representing each word with its context information, where we train our network to minimize the  ...  It is important to learn joint embedding for 3D shapes and text in different shape understanding tasks, such as shape-text matching, retrieval, and shape captioning.  ...  For the point cloud segmentation network, 2500 points are randomly sampled from point clouds with 10000 points to represent a shape.  ... 
arXiv:2107.01872v1 fatcat:c35e3atc7bahrhdl76tfjsvpoe

SeqXY2SeqZ: Structure Learning for 3D Shapes by Sequentially Predicting 1D Occupancy Segments From 2D Coordinates [article]

Zhizhong Han, Guanhui Qiao, Yu-Shen Liu, Matthias Zwicker
2020 arXiv   pre-print
State-of-the-art methods show promising results by representing shapes using implicit functions in 3D that are learned using discriminative neural networks.  ...  Structure learning for 3D shapes is vital for 3D computer vision.  ...  Introduction 3D voxel grids are an attractive representation for 3D structure learning because they can represent shapes with arbitrary topology and they are well suited to convolutional neural network  ... 
arXiv:2003.05559v2 fatcat:nm3q7xfk2jcihcvnlgh7p2ruyi

D2 shape distribution and artificial neural networks for 3D objects recognition

Safae Elhoufi, Aicha Majda, Khalid Abbad
2018 International Journal of Engineering & Technology  
In this paper, we propose a 3D object recognition approach, based on the shape distribution D2 and artificial neural networks.  ...  The values of these histograms feed a multi-layer neural network with back- propagation training.  ...  Fig. 2 : 2 Shape distributions facilitate shape matching because they represent 3D models as functions with a common. Fig. 3 : 3 Overview of the shape similarity search system.  ... 
doi:10.14419/ijet.v7i2.13.11620 fatcat:scp4baeblbbjhneberkp4pvtpu

Few-shot Single-view 3D Reconstruction with Memory Prior Contrastive Network [article]

Zhen Xing and Yijiang Chen and Zhixin Ling and Xiangdong Zhou and Yu Xiang
2022 arXiv   pre-print
With the shape memory, a multi-head attention module is proposed to capture different parts of a candidate shape prior and fuse these parts together to guide 3D reconstruction of novel categories.  ...  In this paper, we present a Memory Prior Contrastive Network (MPCN) that can store shape prior knowledge in a few-shot learning based 3D reconstruction framework.  ...  Using voxels to represent a 3D shape is suitable for 3D CNNs.  ... 
arXiv:2208.00183v1 fatcat:33pwsqtzvbck3oi4xigefmtwli

3D-GMNet: Single-View 3D Shape Recovery as A Gaussian Mixture [article]

Kohei Yamashita, Shohei Nobuhara, Ko Nishino
2020 arXiv   pre-print
In this paper, we introduce 3D-GMNet, a deep neural network for 3D object shape reconstruction from a single image. As the name suggests, 3D-GMNet recovers 3D shape as a Gaussian mixture.  ...  In contrast to voxels, point clouds, or meshes, a Gaussian mixture representation provides an analytical expression with a small memory footprint while accurately representing the target 3D shape.  ...  [18] train the network as a nonlinear function representing the occupancy probability of 3D object shape.  ... 
arXiv:1912.04663v2 fatcat:6uwwkzup3zfwjjcsmq2jferpsy

3D Shape Classification using a Single View

Bo Ding, Lei Tang, Zheng Gao, Yongjun He
2020 IEEE Access  
In order to solve these problems, we propose a novel 3D shape classification method based on Convolutional Neural Network (CNN).  ...  In the recognition stage, the weighted fusion of image clarity evaluation functions is used to select the most representative view for the 3D shape recognition.  ...  The six 3D shape classifiers share the same feature extraction network, but have different classification networks.  ... 
doi:10.1109/access.2020.3035583 fatcat:b36oe6h3bvbj5bxcxskdfpb74y

MPAN: Multi-part Attention Network for Point Cloud Based 3D Shape Retrieval

Zirui Li, Junyu Xu, Yue Zhao, Wenhui Li, Weizhi Nie
2020 IEEE Access  
Meanwhile, by considering the structural relevance of them, the redundancy for representing 3D shapes is removed while the effective information is utilized.  ...  With the development of deep learning technology, great progress has been made in recent years and lots of methods have achieved promising 3D shape retrieval results.  ...  Concretely, it contains a large amount of 3D shapes from 16 shape categories and is split into a training set with 13998 shapes and a testing set with 3894 shapes. B.  ... 
doi:10.1109/access.2020.3018696 fatcat:j43mke27fffwbbv53aun76hpqi

3D-A-Nets: 3D Deep Dense Descriptor for Volumetric Shapes with Adversarial Networks [article]

Mengwei Ren, Liang Niu, Yi Fang
2017 arXiv   pre-print
More specifically, the generator network produces 3D shape features that encourages the clustering of samples from the same category with correct class label, whereas the discriminator network discourages  ...  In this paper, powered with a novel design of adversarial networks (3D-A-Nets), we have developed a novel 3D deep dense shape descriptor (3D-DDSD) to address the challenging issues of efficient and effective  ...  CNN), recurrent neural network (RNN) and an adversarial discriminator for the robust 3D-DDSD for volumetric shapes.  ... 
arXiv:1711.10108v1 fatcat:7kctixmsuffq5avgdrfeyw6sr4

An Improved 3D Shape Recognition Method Based on Panoramic View

Qiang Zheng, Jian Sun, Le Zhang, Wei Chen, Huanhuan Fan
2018 Mathematical Problems in Engineering  
The experimental results show that our approach outperforms DeepPano and can deal with more complex 3D shape recognition problems with a higher diversity of target orientation.  ...  With the development of 2.5D depth sensors, shape recognition is becoming more important in practical applications.  ...  [7] transformed each 3D shape into a 3D grid, and volumetric models were applied for 3D shape recognition with the use of a convolutional deep belief network.  ... 
doi:10.1155/2018/6467957 fatcat:ljurgarxqnashgzftrevl6yvgi

Local Deep Implicit Functions for 3D Shape

Kyle Genova, Forrester Cole, Avneesh Sud, Aaron Sarna, Thomas Funkhouser
2020 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)  
right) whose contributions to an implicit surface reconstruction (middle) are represented by latent vectors decoded by a deep network.  ...  This paper introduces Local Deep Implicit Functions, a 3D shape representation that decomposes an input shape (mesh on left in every triplet) into a structured set of shape elements (colored ellipses on  ...  Like a typical deep implicit function, our LDIF represents a 3D shape as an isocontour of an implicit function decoded with a deep network conditioned on predicted latent variables.  ... 
doi:10.1109/cvpr42600.2020.00491 dblp:conf/cvpr/GenovaCSSF20 fatcat:v3lf6cbybvdbpk72csobi5zpo4
« Previous Showing results 1 — 15 out of 328,325 results