SALA: Soft Assignment Local Aggregation for Parameter Efficient 3D Semantic Segmentation
This work addresses the problem of parameter efficiency in 3D semantic segmentation for researchers and practitioners, offering a significant reduction in model size while maintaining or improving performance.
This paper introduces SALA, a novel point local aggregation function for 3D point cloud semantic segmentation that uses learnable neighbor-to-grid soft assignment. This approach achieves state-of-the-art performance on S3DIS with at least 10x fewer parameters than the previous best method, while also showing competitive results on ScanNet and PartNet.
In this work, we focus on designing a point local aggregation function that yields parameter efficient networks for 3D point cloud semantic segmentation. We explore the idea of using learnable neighbor-to-grid soft assignment in grid-based aggregation functions. Previous methods in literature operate on a predefined geometric grid such as local volume partitions or irregular kernel points. A more general alternative is to allow the network to learn an assignment function that best suits the end task. Since it is learnable, this mapping is allowed to be different per layer instead of being applied uniformly throughout the depth of the network. By endowing the network with the flexibility to learn its own neighbor-to-grid assignment, we arrive at parameter efficient models that achieve state-of-the-art (SOTA) performance on S3DIS with at least 10$\times$ less parameters than the current reigning method. We also demonstrate competitive performance on ScanNet and PartNet compared with much larger SOTA models.