PointResNet: Residual Network for 3D Point Cloud Segmentation and Classification
This addresses a problem in 3D computer vision for applications like augmented reality and robotics, but it appears incremental as it builds on existing residual network concepts.
The paper tackles the challenge of processing irregular 3D point clouds for segmentation and classification by proposing PointResNet, a residual block-based model that directly processes points, achieving the best results for segmentation and comparable results for classification compared to conventional baselines.
Point cloud segmentation and classification are some of the primary tasks in 3D computer vision with applications ranging from augmented reality to robotics. However, processing point clouds using deep learning-based algorithms is quite challenging due to the irregular point formats. Voxelization or 3D grid-based representation are different ways of applying deep neural networks to this problem. In this paper, we propose PointResNet, a residual block-based approach. Our model directly processes the 3D points, using a deep neural network for the segmentation and classification tasks. The main components of the architecture are: 1) residual blocks and 2) multi-layered perceptron (MLP). We show that it preserves profound features and structural information, which are useful for segmentation and classification tasks. The experimental evaluations demonstrate that the proposed model produces the best results for segmentation and comparable results for classification in comparison to the conventional baselines.