Dynamic Graph CNN for Learning on Point Clouds

Wang, Yue; Sun, Yongbin; Liu, Ziwei; Sarma, Sanjay E.; Bronstein, Michael M.; Solomon, Justin

doi:10.1145/3326362

Cited by 4,603 publications

(3,969 citation statements)

References 64 publications

Supporting

Mentioning

3,603

Contrasting

Unclassified

Order By: Relevance

“…Learning-based approaches, and especially those based on deep learning, have recently been proposed specifically to handle point cloud data. The seminal PointNet architecture [19] has inspired a large number of extensions and follow-up works, notably including PointNet++ [20] and Dynamic Graph CNNs [23] for shape classification and segmentation, and more recently PCP-Net [7] for normal and curvature estimation, PU-Net [30] for point cloud upsampling, and PCN for shape completion [31] among many others.…”

Section: Related Workmentioning

confidence: 99%

Effective Rotation-Invariant Point CNN with Spherical Harmonics Kernels

Poulenard

Rakotosaona

Ponty

et al. 2019

2019 International Conference on 3D Vision (3DV)

View full text Add to dashboard Cite

We present a novel rotation invariant architecture operating directly on point cloud data. We demonstrate how rotation invariance can be injected into a recently proposed point-based PCNN architecture, at all layers of the network, achieving invariance to both global shape transformations, and to local rotations on the level of patches or parts, useful when dealing with non-rigid objects. We achieve this by employing a spherical harmonics based kernel at different layers of the network, which is guaranteed to be invariant to rigid motions. We also introduce a more efficient pooling operation for PCNN using space-partitioning datastructures. This results in a flexible, simple and efficient architecture that achieves accurate results on challenging shape analysis tasks including classification and segmentation, without requiring data-augmentation, typically employed by non-invariant approaches.

show abstract

Section: Related Workmentioning

confidence: 99%

Effective Rotation-Invariant Point CNN with Spherical Harmonics Kernels

Poulenard

Rakotosaona

Ponty

et al. 2019

2019 International Conference on 3D Vision (3DV)

View full text Add to dashboard Cite

show abstract

“…With the obtained global feature g, the key is how to acquire the feature for each point. There are two options, one is to duplicate the global feature with N times as in Wang et al (2018), the other is to perform upsampling by the interpolation layer Qi et al (2017b). In the shape segmentation module, two interpolation layers are equipped in our network, which propagate the features from shape level to point level by upsampling.…”

Section: Expansion For Shape Segmentationmentioning

confidence: 99%

LRC-Net: Learning discriminative features on point clouds by encoding local region contexts

Liu

Han

Hong

et al. 2020

Computer Aided Geometric Design

View full text Add to dashboard Cite

Learning discriminative feature directly on point clouds is still challenging in the understanding of 3D shapes. Recent methods usually partition point clouds into local region sets, and then extract the local region features with fixed-size CNN or MLP, and finally aggregate all individual local features into a global feature using simple max pooling. However, due to the irregularity and sparsity in sampled point clouds, it is hard to encode the fine-grained geometry of local regions and their spatial relationships when only using the fixed-size filters and individual local feature integration, which limit the ability to learn discriminative features. To address this issue, we present a novel Local-Region-Context Network (LRC-Net), to learn discriminative features on point clouds by encoding the fine-grained contexts inside and among local regions simultaneously. LRC-Net consists of two main modules. The first module, named intra-region context encoding, is designed for capturing the geometric correlation inside each local region by novel variable-size convolution filter. The second module, named inter-region context encoding, is proposed for integrating the spatial relationships among local regions based on spatial similarity measures. Experimental results show that LRC-Net is competitive with state-of-the-art methods in shape classification and shape segmentation applications.

show abstract

“…The resulting node and edge attribute embeddings are passed through N residual edge convolution and aggregation (ETA) blocks. In the convolution step, we update the edge attributes using a modified version of the edge convolution layer (27), which takes as input a concatenation of node and edge attributes and returns an update to the edge attributes. In our work, g is a multi-layer perceptron, although other neural network architectures are possible.…”

Section: Network Architecturementioning

confidence: 99%

Fast and flexible design of novel proteins using graph neural networks

Strokach¹,

Becerra

Corbi‐Verge

et al. 2019

Preprint

View full text Add to dashboard Cite

Protein structure and function is determined by the arrangement of the linear sequence of amino acids in 3D space. Despite substantial advances, precisely designing sequences that fold into a predetermined shape (the "protein design" problem) remains difficult. We show that a deep graph neural network, ProteinSolver, can solve protein design by phrasing it as a constraint satisfaction problem (CSP). To sidestep the considerable issue of optimizing the network architecture, we first develop a network that is accurately able to solve the related and straightforward problem of Sudoku puzzles. Recognizing that each protein design CSP has many solutions, we train this network on millions of real protein sequences corresponding to thousands of protein structures. We show that our method rapidly designs novel protein sequences and perform a variety of in silico and in vitro validations suggesting that our designed proteins adopt the predetermined structures.One Sentence Summary: A neural network optimized using Sudoku puzzles designs protein sequences that adopt predetermined structures.

show abstract

Dynamic Graph CNN for Learning on Point Clouds

Cited by 4,603 publications

References 64 publications

Effective Rotation-Invariant Point CNN with Spherical Harmonics Kernels

Effective Rotation-Invariant Point CNN with Spherical Harmonics Kernels

LRC-Net: Learning discriminative features on point clouds by encoding local region contexts

Fast and flexible design of novel proteins using graph neural networks

Contact Info

Product

Resources

About