The task of point cloud upsampling aims to acquire dense and uniform point sets from sparse and irregular point sets.
According to our experiments under this fine-grained dataset, we find that state-of-the-art methods are significantly limited by the small variance among subcategories in the same category.
However, due to the irregularity and sparsity in sampled point clouds, it is hard to encode the fine-grained geometry of local regions and their spatial relationships when only using the fixed-size filters and individual local feature integration, which limit the ability to learn discriminative features.
Compared to the previous capsule network based methods, the feature routing on the spatial-aware capsules can learn more discriminative spatial relationships among local regions for point clouds, which establishes a direct mapping between log priors and the spatial locations through feature clusters.
Specifically, L2G-AE employs an encoder to encode the geometry information of multiple scales in a local region at the same time.
In contrast, we propose a deep neural network, called Parts4Feature, to learn 3D global features from part-level information in multiple views.
However, it is hard to capture fine-grained contextual information in hand-crafted or explicit manners, such as the correlation between different areas in a local region, which limits the discriminative ability of learned features.
Ranked #35 on 3D Part Segmentation on ShapeNet-Part