Filters








75 Hits in 6.8 sec

O-CNN

Peng-Shuai Wang, Yang Liu, Yu-Xiao Guo, Chun-Yu Sun, Xin Tong
2017 ACM Transactions on Graphics  
We present O-CNN, an Octree-based Convolutional Neural Network (CNN) for 3D shape analysis.  ...  O-CNN supports various CNN structures and works for 3D shapes in different representations.  ...  ACKNOWLEDGMENTS We wish to thank the authors of [Chang et al. 2015; Wu et al. 2015] for sharing their 3D model datasets with the public, the authors of [Qi et al. 2017; Yi et al. 2016] for providing  ... 
doi:10.1145/3072959.3073608 fatcat:jvviegjchzalpawfrkd6hyyi24

Adaptive O-CNN: A Patch-based Deep Representation of 3D Shapes [article]

Peng-Shuai Wang and Chun-Yu Sun and Yang Liu and Xin Tong
2018 arXiv   pre-print
We present an Adaptive Octree-based Convolutional Neural Network (Adaptive O-CNN) for efficient 3D shape encoding and decoding.  ...  Based on this adaptive patch-based representation, we propose an Adaptive O-CNN encoder and decoder for encoding and decoding 3D shapes.  ...  ACKNOWLEDGMENTS We wish to thank the authors of ModelNet and ShapeNet for sharing data, Stephen Lin for proofreading the paper, and the anonymous reviewers for their valuable feedback.  ... 
arXiv:1809.07917v1 fatcat:ted5t6qtffbm3agjz7nedc6lby

RocNet: Recursive Octree Network for Efficient 3D Deep Representation [article]

Juncheng Liu, Steven Mills, Brendan McCane
2020 arXiv   pre-print
We introduce a deep recursive octree network for the compression of 3D voxel data. Our network compresses a voxel grid of any size down to a very small latent space in an autoencoder-like network.  ...  We demonstrate the effectiveness and efficiency of our proposed method on several publicly available datasets with three experiments: 3D shape classification, 3D shape reconstruction, and shape generation  ...  O-CNN [1] uses an octree-based 3D convolutional network by using fast shuffled key and hash table search.  ... 
arXiv:2008.03875v1 fatcat:l6okennx3zha7maiyd6dhhmxz4

Deep Octree-based CNNs with Output-Guided Skip Connections for 3D Shape and Scene Completion

Peng-Shuai Wang, Yang Liu, Xin Tong
2020 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)  
Our network is built upon the octree-based CNNs (O-CNN) with U-Net like structures, which enjoys high computational and memory efficiency and supports to construct a very deep network structure for 3D  ...  We show that with these simple adaptionsoutput-guided skip-connection and deeper O-CNN (up to 70 layers), our network achieves state-of-the-art results in 3D shape completion and semantic scene computation  ...  Neural networks have been extensively used for shape completion.  ... 
doi:10.1109/cvprw50498.2020.00141 dblp:conf/cvpr/Wang0020 fatcat:zbb4pg6dzbdjtff3b2bt62ismi

Deep Octree-based CNNs with Output-Guided Skip Connections for 3D Shape and Scene Completion [article]

Peng-Shuai Wang and Yang Liu and Xin Tong
2020 arXiv   pre-print
Our network is built upon the octree-based CNNs (O-CNN) with U-Net like structures, which enjoys high computational and memory efficiency and supports to construct a very deep network structure for 3D  ...  We show that with these simple adaptions -- output-guided skip-connection and deeper O-CNN (up to 70 layers), our network achieves state-of-the-art results in 3D shape completion and semantic scene computation  ...  Neural networks have been extensively used for shape completion.  ... 
arXiv:2006.03762v1 fatcat:sjakew3nl5d7xmeol4vrskcejy

Deep Implicit Moving Least-Squares Functions for 3D Reconstruction [article]

Shi-Lin Liu, Hao-Xiang Guo, Hao Pan, Peng-Shuai Wang, Xin Tong, Yang Liu
2021 arXiv   pre-print
Our IMLSNet predicts an octree structure as a scaffold for generating MLS points where needed and characterizes shape geometry with learned local priors.  ...  However, their discrete nature prevents them from representing continuous and fine geometry, posing a major issue for learning-based shape generation.  ...  Scaffold prediction We use the octree-based convolutional neural network (O-CNN) autoencoder [49, 51] to generate the scaffold.  ... 
arXiv:2103.12266v2 fatcat:mkb4xs2lijadtfmp572al6yj6m

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
In recent years, 3D computer vision and Geometry Deep Learning gain more and more attention. Many advanced techniques for 3D shapes have been proposed for different applications.  ...  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  ...  [105] also proposed a convolutional neural network based on octree called O-CNN, where the model also removes pointers like shallow octree [80] and stores the octree data and structure by a series  ... 
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  
In recent years, 3D computer vision and geometry deep learning have gained ever more attention. Many advanced techniques for 3D shapes have been proposed for different applications.  ...  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  ...  [22] also proposed an octree-based convolutional neural network called O-CNN, where the model also removes pointers like a shallow octree [21] and stores the octree data and structure using a series  ... 
doi:10.1007/s41095-020-0174-8 fatcat:kpoynaixq5esbek63bovybisfa

Interpolation-Aware Padding for 3D Sparse Convolutional Neural Networks [article]

Yu-Qi Yang, Peng-Shuai Wang, Yang Liu
2021 arXiv   pre-print
Sparse voxel-based 3D convolutional neural networks (CNNs) are widely used for various 3D vision tasks.  ...  Sparse voxel-based 3D CNNs create sparse non-empty voxels from the 3D input and perform 3D convolution operations on them only.  ...  [5] bring the recurrent neural network to voxel-based 3D decoders for inferring 3D shapes from multiview images.  ... 
arXiv:2108.06925v1 fatcat:5rgz3j2r4nga5o7bquj7qbzz7q

Image-based 3D Object Reconstruction: State-of-the-Art and Trends in the Deep Learning Era

Xianfeng Han, Hamid Laga, Mohammed Bennamoun
2019 IEEE Transactions on Pattern Analysis and Machine Intelligence  
Since 2015, image-based 3D reconstruction using convolutional neural networks (CNN) has attracted increasing interest and demonstrated an impressive performance.  ...  We organize the literature based on the shape representations, the network architectures, and the training mechanisms they use.  ...  O-CNN supports various CNN structures and works with 3D shapes of different representations.  ... 
doi:10.1109/tpami.2019.2954885 pmid:31751229 fatcat:hc76yes6avdy5byyy7flovj5wa

A Review on Deep Learning Approaches for 3D Data Representations in Retrieval and Classifications

Abubakar Sulaiman Gezawa, Yan Zhang, Qicong Wang, Lei Yunqi
2020 IEEE Access  
Therefore, it can be concluded that deep learning together with a suitable 3D data representation gives an effective approach for improving the performance of 3D shape analysis.  ...  Deep learning approach has been used extensively in image analysis tasks.  ...  [24] proposed O-CNN which is an Octree-based Convolutional Neural Network (O-CNN) for object classification, retrieval and segmentation tasks.  ... 
doi:10.1109/access.2020.2982196 fatcat:jnya5rscynf3zm7efuucqxafri

PointGrid: A Deep Network for 3D Shape Understanding

Truc Le, Ye Duan
2018 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition  
In this work, we propose the PointGrid, a 3D convolutional network that incorporates a constant number of points within each grid cell thus allowing the network to learn higher order local approximation  ...  Volumetric grid is widely used for 3D deep learning due to its regularity.  ...  [55] for making PointNet and O-CNN publicly available.  ... 
doi:10.1109/cvpr.2018.00959 dblp:conf/cvpr/LeD18 fatcat:r7pscovahzemvfpqnynhy2lnoq

Quadtree Convolutional Neural Networks [chapter]

Pradeep Kumar Jayaraman, Jianhan Mei, Jianfei Cai, Jianmin Zheng
2018 Lecture Notes in Computer Science  
This paper presents a Quadtree Convolutional Neural Network (QCNN) for efficiently learning from image datasets representing sparse data such as handwriting, pen strokes, freehand sketches, etc.  ...  We study QCNN on four sparse image datasets for sketch classification and simplification tasks.  ...  Acknowledgements We thank the anonymous reviewers for their constructive comments. This research is supported by the National Research Foundation under Virtual Singapore Award No.  ... 
doi:10.1007/978-3-030-01231-1_34 fatcat:gynykn7lc5bo5pntbs2qcgi34m

SO-Net: Self-Organizing Network for Point Cloud Analysis [article]

Jiaxin Li, Ben M. Chen, Gim Hee Lee
2018 arXiv   pre-print
In recognition tasks such as point cloud reconstruction, classification, object part segmentation and shape retrieval, our proposed network demonstrates performance that is similar with or better than  ...  This paper presents SO-Net, a permutation invariant architecture for deep learning with orderless point clouds.  ...  Introduction After many years of intensive research, convolutional neural networks (ConvNets) is now the foundation for many state-of-the-art computer vision algorithms, e.g. image recognition, object  ... 
arXiv:1803.04249v4 fatcat:fxqei6t5nfdt7eax6puvljlivu

VoxSegNet: Volumetric CNNs for Semantic Part Segmentation of 3D Shapes [article]

Zongji Wang, Feng Lu
2018 arXiv   pre-print
Voxel is an important format to represent geometric data, which has been widely used for 3D deep learning in shape analysis due to its generalization ability and regular data format.  ...  In this paper, we propose a novel volumetric convolutional neural network, which could extract discriminative features encoding detailed information from voxelized 3D data under a limited resolution.  ...  The point clouds were mapped to grid-octree structure and then fed into a U-shaped encoder-decoder convolutional neural network to predict the label for each voxel. Wang et al.  ... 
arXiv:1809.00226v1 fatcat:rlxoahgzdvdcbbdmv3j453tp2q
« Previous Showing results 1 — 15 out of 75 results