1 Abstract Due to recent technological advances, 3D geometry is becoming commonplace on all the entry-level PCs connected to the Internet, and multimedia-enabled systems must be prepared to a great increase of complex 3D content in the future. However, while Computer Graphics research has solved many problems related to the creation and manipulation of digital 3D shapes, work on how to extract,...

Face alignment and 3D face reconstruction are traditionally accomplished as separated tasks. By exploring the strong correlation between 2D landmarks and 3D shapes, in contrast, we propose a joint face alignment and 3D face reconstruction method to simultaneously solve these two problems for 2D face images of arbitrary poses and expressions. This method, based on a summation model of 3D face sh...

This paper reviews main approaches to 3D shape perception in both human and computer vision. The approaches are evaluated with respect to their plausibility of generating adequate explanations of human vision. The criterion for plausibility is provided by existing psychophysical results. A new theory of 3D shape perception is then outlined. According to this theory, human perception of shapes c...

The research described in this thesis has focused on the composition of novel 3D shapes (e.g., depth maps and triangulated meshes) by reusing parts of existing shapes. Specifically, two methods were designed [28, 27]. The first is a novel tool for easy composition of existing 3D models (i.e., triangulated meshes) into novel models [28]. This tool was designed with the intention of providing a m...

Matching 3D shapes is a fundamental problem in Medical Imaging with many applications including, but not limited to, shape deformation analysis, tracking etc. Matching 3D shapes poses a computationally challenging task. The problem is especially hard when the transformation sought is diffeomorphic and non-rigid between the shapes being matched. In this paper, we propose a novel and computationa...

Assessing the similarity among 3D shapes is a challenging research topic, and effective shape descriptions have to be devised in order to support the matching process. There is a growing consensus that shapes are recognized and coded mentally in terms of relevant parts and their spatial configuration, or structure. The presentation will discuss the definition and use of structural descriptions ...

Motivation: If two objects can be set in correspondence, then a class of objects can be modeled in the following way. First all objects are set in correspondence, one by one, with an base object. These correspondences represent warpings of one object to other objects in the same class of instances. When these warpings are combined in a linear network as described in [5], we have then created a ...

We present a bio-inspired calculus for describing 3D shapes moving in a space. A shape forms a 3D process when combined with a behaviour. Behaviours are speci ed with a timed CCS-like process algebra using a notion of channel to naturally model binding sites on the surface of shapes. The calculus embeds collision detection and response, binding of compatible 3D processes and split of composed 3D

Rapidly increasing 3D shape application has led to the development of content-based 3D shape retrieval research. In this paper, we proposed a new retrieval method. The method is constructed on a spatial distribution computation of sampling points on the surface of 3D shape. The contribution is that we use an inner cylinder to contain the points distributed nearer on the largest principal axis, ...

In this paper, we propose a highly efficient metric learning approach to non-rigid 3D shape analysis. From a training set of 3D shapes from different classes, we learn a transformation of the shapes which optimally enforces a clustering of shapes from the same class. In contrast to existing approaches, we do not perform a transformation of individual local point descriptors, but a linear embedd...