3D Graph Neural Networks for RGBD Semantic Segmentation

By Xiaojuan Qi, Renjie Liao, Jiaya Jia, Sanja Fidler, Raquel Urtasun. IEEE International Conference on Computer Vision, 2017

475 citations recorded by the source corpus.

Subfields: Computational Mechanics, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/4ebf5a9c35f40d3343a2376d0b75d8d7a3126d8d

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computational Mechanics

In Computer Vision and Pattern Recognition

Citations in the corpus

Efficient Interactive Annotation of Segmentation Datasets with Polygon-RNN++ (2018, methodology): “GNN has been proven efficient for semantic segmentation [24], where it was used at pixel-level.”
Learning Depth with Convolutional Spatial Propagation Network (2018, background): “It also serves as a foundation to support other computer vision problems, such as 3D reconstruction [5], [6] and recognition [7], [8].”
LanczosNet: Multi-Scale Deep Graph Convolutional Networks (2019, background): “Moreover, many applications have recently emerged for GNNs, including community detection [41], situation recognition [42], RGBD semantic segmentation [43], few-shot learning [21], probabilistic inference [44], continuous control of reinforcement learning [45, 46] and so on.”
Point-GNN: Graph Neural Network for 3D Object Detection in a Point Cloud (2020, methodology): “[15] uses a recurrent GNN for the semantic segmentation on RGBD data.”
PPF-FoldNet: Unsupervised Learning of Rotation Invariant 3D Local Descriptors (2018, unknown): “[32] apply GCNs to RGB-D semantic segmentation.”
Deep Parametric Continuous Convolutional Neural Networks (2018, background): “In [23], GGNNs are applied to point cloud segmentation, achieving significant improvements over the state-of-the-art.”
Skeleton-Based Action Recognition with Spatial Reasoning and Temporal Stack Learning (2018, methodology): “[22] use 3D graph neural networks for RGBD semantic segmentation.”
Depth-aware CNN for RGB-D Segmentation (2018, methodology): “Following [4], global pooling is used to compute a C -dim vector from conv 7.”
Bi-directional Cross-Modality Feature Propagation with Separation-and-Aggregation Gate for RGB-D Semantic Segmentation (2020, unknown): “formation. Besides, we can see that utilizing depth data as extra supervision (such as [44,39]) could make network more robust than general RGB-D methods that take both RGB and depth as input sources [26,6,27]. However, our results demonstrate that once the input RGB-D information could be eectively”
JSIS3D: Joint Semantic-Instance Segmentation of 3D Point Clouds With Multi-Task Pointwise Networks and Multi-Value Conditional Random Fields (2019, unknown): “, [41, 14, 20, 46, 42, 27, 34].”
Cylindrical and Asymmetrical 3D Convolution Networks for LiDAR Segmentation (2020, background): “Mainstream methods [25, 31, 39, 36, 32, 20, 10, 45, 41, 33, 24, 26] of indoor point cloud segmentation learn the point features based on the raw point directly, which are often based on the pioneering work, i.”
Pattern-Affinitive Propagation Across Depth, Surface Normal and Semantic Segmentation (2019, unknown): “[45] designed a 3D graph neural network to fuse the depth information for segmentation.”
Tangent Convolutions for Dense Prediction in 3D (2018, background): “Other applications of deep networks consider RGB-D images, which can be treated with fully-convolutional networks [16, 29, 36] and graph neural networks [41].”
DeepSphere: Efficient spherical Convolutional Neural Network with HEALPix sampling for cosmological applications (2018, methodology): “Thanks to their versatility, graph neural networks have been used in a variety of tasks, such as identifying diseases from brain connectivity networks [61] or population graphs [62], designing drugs using molecular graphs [63], segmenting 3D point clouds [64], optimizing shapes to be aerodynamic [65”
Synthetic Data for Deep Learning (2019, methodology): “The original paper [557] presented state of the art results in semantic scene completion, but, naturally, SUNCG has been used for many different tasks related to depth estimation, indoor navigation, SLAM, and others [2,108,231,385,465], and it often serves as the basis for scene understanding compet”