PointAtrousGraph: Deep Hierarchical Encoder-Decoder with Point Atrous Convolution for Unorganized 3D Points

Abstract

Motivated by the success of encoding multi-scale contextual information for image analysis, we propose our PointAtrousGraph (PAG) - a deep permutation-invariant hierarchical encoder-decoder for efficiently exploiting multi-scale edge features in point clouds. Our PAG is constructed by several novel modules, such as Point Atrous Convolution (PAC), Edge-preserved Pooling (EP) and Edge-preserved Unpooling (EU). Similar with atrous convolution, our PAC can effectively enlarge receptive fields of filters and thus densely learn multi-scale point features. Following the idea of non-overlapping max-pooling operations, we propose our EP to preserve critical edge features during subsampling. Correspondingly, our EU modules gradually recover spatial information for edge features. In addition, we introduce chained skip subsampling/upsampling modules that directly propagate edge features to the final stage. Particularly, our proposed auxiliary loss functions can further improve our performance. Experimental results show that our PAG outperform previous state-of-the-art methods on various 3D semantic perception applications.

11 pages, 10 figures

References in corpus (6)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Rethinking Atrous Convolution for Semantic Image Segmentation
• DSSD : Deconvolutional Single Shot Detector
• InfoVAE: Information Maximizing Variational Autoencoders
• PointSIFT: A SIFT-like Network Module for 3D Point Cloud Semantic Segmentation
• Beyond Skip Connections: Top-Down Modulation for Object Detection

Cited by in corpus (2)

• Deep Learning for 3D Point Clouds: A Survey
• Are We Hungry for 3D LiDAR Data for Semantic Segmentation? A Survey and Experimental Study