Learning Gaussian Instance Segmentation in Point Clouds

Abstract

This paper presents a novel method for instance segmentation of 3D point clouds. The proposed method is called Gaussian Instance Center Network (GICN), which can approximate the distributions of instance centers scattered in the whole scene as Gaussian center heatmaps. Based on the predicted heatmaps, a small number of center candidates can be easily selected for the subsequent predictions with efficiency, including i) predicting the instance size of each center to decide a range for extracting features, ii) generating bounding boxes for centers, and iii) producing the final instance masks. GICN is a single-stage, anchor-free, and end-to-end architecture that is easy to train and efficient to perform inference. Benefited from the center-dictated mechanism with adaptive instance size selection, our method achieves state-of-the-art performance in the task of 3D instance segmentation on ScanNet and S3DIS datasets.

References in corpus (10)

• Focal Loss for Dense Object Detection
• ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenes
• Semantic Instance Segmentation with a Discriminative Loss Function
• Path Aggregation Network for Instance Segmentation
• Learning Object Bounding Boxes for 3D Instance Segmentation on Point Clouds
• Weakly Supervised Learning of Instance Segmentation with Inter-pixel Relations
• 3D-BEVIS: Bird's-Eye-View Instance Segmentation
• Associatively Segmenting Instances and Semantics in Point Clouds
• Semi-convolutional Operators for Instance Segmentation
• Rethinking Task and Metrics of Instance Segmentation on 3D Point Clouds

Cited by in corpus (4)

• Advancements in Point Cloud-Based 3D Defect Detection and Classification for Industrial Systems: A Comprehensive Survey
• MaskGroup: Hierarchical Point Grouping and Masking for 3D Instance Segmentation
• PIG-Net: Inception based Deep Learning Architecture for 3D Point Cloud Segmentation
• Point Cloud Instance Segmentation with Semi-supervised Bounding-Box Mining