NeuSDFusion: A Spatial-Aware Generative Model for 3D Shape Completion, Reconstruction, and Generation
This addresses the need for more diverse and spatially coherent 3D shapes in computer graphics and AI, representing a novel method for a known bottleneck rather than an incremental improvement.
The paper tackles the problem of limited versatility and spatial inconsistency in 3D shape generation by introducing a spatial-aware framework that uses 2D plane representations and a transformer-based autoencoder, resulting in consistent outperformance of state-of-the-art methods across tasks like unconditional generation and text-to-shape synthesis.
3D shape generation aims to produce innovative 3D content adhering to specific conditions and constraints. Existing methods often decompose 3D shapes into a sequence of localized components, treating each element in isolation without considering spatial consistency. As a result, these approaches exhibit limited versatility in 3D data representation and shape generation, hindering their ability to generate highly diverse 3D shapes that comply with the specified constraints. In this paper, we introduce a novel spatial-aware 3D shape generation framework that leverages 2D plane representations for enhanced 3D shape modeling. To ensure spatial coherence and reduce memory usage, we incorporate a hybrid shape representation technique that directly learns a continuous signed distance field representation of the 3D shape using orthogonal 2D planes. Additionally, we meticulously enforce spatial correspondences across distinct planes using a transformer-based autoencoder structure, promoting the preservation of spatial relationships in the generated 3D shapes. This yields an algorithm that consistently outperforms state-of-the-art 3D shape generation methods on various tasks, including unconditional shape generation, multi-modal shape completion, single-view reconstruction, and text-to-shape synthesis. Our project page is available at https://weizheliu.github.io/NeuSDFusion/ .