PointGrow: Autoregressively Learned Point Cloud Generation with Self-Attention
This work addresses the challenge of 3D point cloud generation for applications like unsupervised feature learning and shape arithmetic, but it appears incremental as it builds on autoregressive models with self-attention enhancements.
The authors tackled the problem of generating diverse and realistic 3D point clouds from scratch or with semantic conditioning, achieving satisfying performance in terms of realism and diversity as shown in extensive evaluations.
Generating 3D point clouds is challenging yet highly desired. This work presents a novel autoregressive model, PointGrow, which can generate diverse and realistic point cloud samples from scratch or conditioned on semantic contexts. This model operates recurrently, with each point sampled according to a conditional distribution given its previously-generated points, allowing inter-point correlations to be well-exploited and 3D shape generative processes to be better interpreted. Since point cloud object shapes are typically encoded by long-range dependencies, we augment our model with dedicated self-attention modules to capture such relations. Extensive evaluations show that PointGrow achieves satisfying performance on both unconditional and conditional point cloud generation tasks, with respect to realism and diversity. Several important applications, such as unsupervised feature learning and shape arithmetic operations, are also demonstrated.