Colloquium Details

Heat-Driven Approaches for 3D Shape Analysis

Speaker: Yi Fang, Purdue

Location: Warren Weaver Hall 1302

Date: March 21, 2012, 11:30 a.m.

Host: Denis Zorin


Recent development in technology for digital acquisition of 3D models has led to a rapid growth of the number of available three dimensional (3D) objects across areas as diverse as engineering, medicine and biology. Within these diverse areas, an intrinsic geometric description is often necessary to compare, characterize, and interpret these objects. The effective and efficient interpretation of 3D models is often challenging due to the non-rigidity of the shape, the presence of geometric noise and the large volume of the shape database. My research is concentrated on the development of novel heat-driven approaches based on Heat Kernel for efficient 3D shape analysis including shape matching, shape segmentation and shape retrieving. The novelty of heat-driven approaches is derived from an analogy between the process of 3D shape interpretation and that of heat transfer, in which all points on the surface contribute simultaneously and globally to reveal intrinsic similarities between regions of shape, meaningful coordinates for embedding the shape, and exemplar points that lie at optimal positions for heat transfer. The approaches exploit the intelligence of the heat as a global structure-aware message on a meshed surface, and are capable of exploring the intrinsic geometric features of the shape. We demonstrate the performance of the heat-driven approaches for effective non-rigid 3D shape registration, robust segmentation of 3D models, and efficient retrieval of 3D models with applications across multi-disciplines. The experimental results indicate that heat-driven approaches are able to reveal human perceptual consistent interpretation of 3D shape in a robust fashion with no reference to prior knowledge. In addition, our proposed heat-driven approaches are very general and have great potential for applications to a broad range of research fields, for example, biological networks, social networks and semantic analysis of documents.


In-person attendance only available to those with active NYU ID cards.

How to Subscribe