PanopticFusion: Online Volumetric Semantic Mapping at the Level of Stuff and Things

Abstract

We propose PanopticFusion, a novel online volumetric semantic mapping system at the level of stuff and things. In contrast to previous semantic mapping systems, PanopticFusion is able to densely predict class labels of a background region (stuff) and individually segment arbitrary foreground objects (things). In addition, our system has the capability to reconstruct a large-scale scene and extract a labeled mesh thanks to its use of a spatially hashed volumetric map representation. Our system first predicts pixel-wise panoptic labels (class labels for stuff regions and instance IDs for thing regions) for incoming RGB frames by fusing 2D semantic and instance segmentation outputs. The predicted panoptic labels are integrated into the volumetric map together with depth measurements while keeping the consistency of the instance IDs, which could vary frame to frame, by referring to the 3D map at that moment. In addition, we construct a fully connected conditional random field (CRF) model with respect to panoptic labels for map regularization. For online CRF inference, we propose a novel unary potential approximation and a map division strategy. We evaluated the performance of our system on the ScanNet (v2) dataset. PanopticFusion outperformed or compared with state-of-the-art offline 3D DNN methods in both semantic and instance segmentation benchmarks. Also, we demonstrate a promising augmented reality application using a 3D panoptic map generated by the proposed system.

8 pages, 6 figures, Accepted to 2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)

References in corpus (8)

• Efficient Inference in Fully Connected CRFs with Gaussian Edge Potentials
• ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenes
• MASC: Multi-scale Affinity with Sparse Convolution for 3D Instance Segmentation
• GSPN: Generative Shape Proposal Network for 3D Instance Segmentation in Point Cloud
• 3D-SIS: 3D Semantic Instance Segmentation of RGB-D Scans
• MID-Fusion: Octree-based Object-Level Multi-Instance Dynamic SLAM
• 3DMV: Joint 3D-Multi-View Prediction for 3D Semantic Scene Segmentation
• TextureNet: Consistent Local Parametrizations for Learning from High-Resolution Signals on Meshes

Cited by in corpus (5)

• PointGroup: Dual-Set Point Grouping for 3D Instance Segmentation
• Spatial Semantic Embedding Network: Fast 3D Instance Segmentation with Deep Metric Learning
• RevealNet: Seeing Behind Objects in RGB-D Scans
• OccuSeg: Occupancy-aware 3D Instance Segmentation
• Panoptic Segmentation: A Review