SEGCloud: Semantic Segmentation of 3D Point Clouds
This addresses the problem of fine-grained semantic labeling for agents in real-world environments, representing an incremental improvement over existing methods.
The paper tackled 3D semantic segmentation of point clouds by proposing SEGCloud, an end-to-end framework that combines neural networks, trilinear interpolation, and fully connected Conditional Random Fields to achieve point-level segmentation with global consistency, resulting in performance comparable or superior to state-of-the-art on four indoor and outdoor datasets.
3D semantic scene labeling is fundamental to agents operating in the real world. In particular, labeling raw 3D point sets from sensors provides fine-grained semantics. Recent works leverage the capabilities of Neural Networks (NNs), but are limited to coarse voxel predictions and do not explicitly enforce global consistency. We present SEGCloud, an end-to-end framework to obtain 3D point-level segmentation that combines the advantages of NNs, trilinear interpolation(TI) and fully connected Conditional Random Fields (FC-CRF). Coarse voxel predictions from a 3D Fully Convolutional NN are transferred back to the raw 3D points via trilinear interpolation. Then the FC-CRF enforces global consistency and provides fine-grained semantics on the points. We implement the latter as a differentiable Recurrent NN to allow joint optimization. We evaluate the framework on two indoor and two outdoor 3D datasets (NYU V2, S3DIS, KITTI, Semantic3D.net), and show performance comparable or superior to the state-of-the-art on all datasets.