Active appearance models (AAMs) have demonstrated great utility when being employed for non-rigid face alignment/tracking. The "simultaneous" algorithm for fitting an AAM achieves good non-rigid face registration performance, but has poor real time performance (2-3 fps). The "project-out" algorithm for fitting an AAM achieves faster than real time performance (> 200 fps) but suffers from poor g...

Non-rigid registration computes an alignment between a source surface with target in non-rigid manner. In the past decade, advances 3D sensing technologies that can measure time-varying surfaces, has been applied for acquisition of deformable shapes and wide range applications. This survey presents comprehensive review methods shapes, focusing on techniques related to dynamic shape reconstructi...

We develop a new algorithm for the pairwise protein structure alignment. Our algorithm has two phases. The first phase aligns all the possible local regions between two proteins. Each local region of a protein consists of a series of consecutive Cα atoms in the backbone of the protein. The second phase derives a global rigid body transformation that combines some aligned local regions to achiev...

We present an approach to find dense point-to-point correspondences between two deformed surfaces corresponding to different postures of the same non-rigid object in a fully automatic way. The approach requires no prior knowledge about the shapes being registered or the initial alignment of the shapes. We consider surfaces that are represented by possibly incomplete triangular meshes. We model ...

In this paper, we propose using a segmented example model to perform a semantic oriented segmentation of rigid 3D models of the same class (tables, chairs, etc.). For this, we introduce an alignment method that maps the meaningful parts of the models and we develop a novel approach based on random walks to transfer a consistent segmentation from the example to the target model. The example-driv...

We present an approach general enough to apply to recognition of complex rigid 3D objects from either a single intensity image or a single range image. Within the general paradigm of recognition by alignment, we address (1) deenition and detection of primitives, (2) indexing to model hypotheses, (3) constructing view sphere models from sensed data, and (4) aligning model and sensed features for...