• Corpus ID: 145043385

Face Tracking with Multilinear ( Tensor ) Active Appearance Models

  title={Face Tracking with Multilinear ( Tensor ) Active Appearance Models},
  author={Weiguang Si and Kota Yamaguchi and M. Alex O. Vasilescu},
Face tracking in an unconstrained environment must contend with images that vary with viewpoint, illumination, expression, identity, and other causal factors. In a statistical approach, the multifactor nature of the image data makes the aforementioned problem amenable to analysis in a multilinear framework. In this paper, we propose Multilinear (Tensor) Active Appearance Models (MAAMs). The MAAM is a multilinear statistical model of facial appearance and shape that generalizes the linear Active… 

Figures from this paper

Compositional Hierarchical Tensor Factorization: Representing Hierarchical Intrinsic and Extrinsic Causal Factors
This paper proposes a unified tensor model of wholes and parts, and introduces a compositional hierarchical tensor factorization that disentangles the hierarchical causal structure of object image formation, and subsumes multilinear block tensor decomposition as a special case.


Bilinear Active Appearance Models
A gradient-descent algorithm is derived to efficiently fit Bilinear Active Appearance Models (BAAMs), an extension of AAMs that effectively decouple changes due to pose and expression/identity and shows how the pose normalization provided by BAAMs increase the recognition performance of commercial systems.
Person-Independent Monocular Tracking of Face and Facial Actions with Multilinear Models
This paper presents a monocular method of tracking faces and facial actions using a multilinear face model that treats interpersonal and intrapersonal shape variations separately and is the key to tracking Faces and facial Actions of arbitrary people in real time with no pre-learned individual face models.
Face transfer with multilinear models
Face Transfer is a method for mapping videorecorded performances of one individual to facial animations of another, based on a multilinear model of 3D face meshes that separably parameterizes the space of geometric variations due to different attributes.
Interpreting face images using active appearance models
A fast, robust method of interpreting face images using an Active Appearance Model (AAM), which contains a statistical model of shape and grey level appearance which can generalise to almost any face.
Multilinear Projection for Appearance-Based Recognition in the Tensor Framework
This work defines the identity tensor and the pseudo-inverse Tensor and employs them to develop a multilinear projection algorithm, which is natural for performing recognition in the tensor algebraic framework.
Multilinear projection for face recognition via canonical decomposition
  • M. Alex O. Vasilescu
  • Computer Science
    2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition - Workshops
  • 2010
A new multilinear projection algorithm for appearance-based recognition in a tensor framework based on the canonical decomposition of tensors is introduced and its results in simultaneously inferring the identity, view, illumination, etc., coefficient vectors of an unlabeled test image are compared.
Multilinear Analysis of Image Ensembles: TensorFaces
This work considers the multilinear analysis of ensembles of facial images that combine several modes, including different facial geometries (people), expressions, head poses, and lighting conditions, and concludes that the resulting "TensorFaces" representation has several advantages over conventional eigenfaces.
Active Appearance Models Revisited
This work proposes an efficient fitting algorithm for AAMs based on the inverse compositional image alignment algorithm and shows that the effects of appearance variation during fitting can be precomputed (“projected out”) using this algorithm and how it can be extended to include a global shape normalising warp.
View-based active appearance models
  • Tim Cootes, K. N. Walker, C. Taylor
  • Computer Science
    Proceedings Fourth IEEE International Conference on Automatic Face and Gesture Recognition (Cat. No. PR00580)
  • 2000
A small number of 2D statistical models are demonstrated to capture the shape and appearance of a face from any viewpoint (full profile to front-to-parallel) and to estimate head pose and to synthesize faces from unseen viewpoints.
Expression Transfer between Photographs through Multilinear AAM's
This work proposes a novel approach for expression transfer based on color images, combining active appearance models and multilinear analysis, and attacks this problem with methods developed by the computer vision community for facial expression analysis and recognition.