SPLATNet: Sparse Lattice Networks for Point Cloud Processing
This work addresses the computational and memory inefficiencies in point cloud processing for applications like 3D segmentation, representing an incremental improvement over existing methods.
The paper tackles the problem of inefficient point cloud processing by introducing SPLATNet, a network architecture that uses sparse bilateral convolutional layers to operate directly on point clouds in a high-dimensional lattice, achieving state-of-the-art performance on 3D segmentation tasks.
We present a network architecture for processing point clouds that directly operates on a collection of points represented as a sparse set of samples in a high-dimensional lattice. Naively applying convolutions on this lattice scales poorly, both in terms of memory and computational cost, as the size of the lattice increases. Instead, our network uses sparse bilateral convolutional layers as building blocks. These layers maintain efficiency by using indexing structures to apply convolutions only on occupied parts of the lattice, and allow flexible specifications of the lattice structure enabling hierarchical and spatially-aware feature learning, as well as joint 2D-3D reasoning. Both point-based and image-based representations can be easily incorporated in a network with such layers and the resulting model can be trained in an end-to-end manner. We present results on 3D segmentation tasks where our approach outperforms existing state-of-the-art techniques.