3D Shape Modeling and Analysis

Computational approaches to 3D shape modeling and analysis are concerned with how machines can represent, reconstruct, and reason about the geometry of physical objects — from raw sensor measurements like point clouds to structured representations like meshes and implicit surfaces. Getting this right matters for applications ranging from robotics and autonomous vehicles to medical imaging and digital fabrication, where accurate geometric understanding directly affects real-world decisions. Deep learning has recently transformed the area, with methods like neural radiance fields enabling photo-realistic reconstruction from sparse image sets and learned mesh segmentation automating tasks that once required manual annotation. Open questions remain around generalizing these models to novel object categories with little training data, and around building representations that are simultaneously efficient to process, easy to deform, and faithful to fine surface detail.

Works

48,716

Total citations

622,113

Keywords

Deep LearningPoint Clouds3D ReconstructionMesh SegmentationShape RepresentationNeural Radiance Fields