3D morphable face models revisited

Patel, Archana; Smith, William A. P.

doi:10.1109/cvprw.2009.5206522

Cited by 47 publications

(55 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Various techniques have been proposed for 3-D facial fitting. Blanz showed 3-D face reconstructions using a 3-D morphable model and optic flow algorithm [1,2,8].…”

Section: Avatar Appearance In Second Lifementioning

confidence: 99%

Automatic 3-D Facial Fitting Technique for a Second Life Avatar

Dohi

Ishizuka

2011

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. This paper describes an automatic 3-D facial fitting technique for a Second Life avatar. It is often difficult to create an original avatar by yourself that resembles a real person. In Second Life, the combinations of systemdefined parameters deform the shape of the avatar, and we can't control each vertex directly. It needs to encode the deformation information into many parameters. As a reference target, we make use of MRI data that scans the real face. It is stable and not affected by lighting and diffused reflection. In our experiments, we picked up 428 vertices on the base model for facial fitting. Using the iteration technique, more than 50% of vertices are just on the reference targets, and more than 85 % are within +/-3mm of errors.

show abstract

“…Various techniques have been proposed for 3-D facial fitting. Blanz showed 3-D face reconstructions using a 3-D morphable model and optic flow algorithm [1,2,8].…”

Section: Avatar Appearance In Second Lifementioning

confidence: 99%

Automatic 3-D Facial Fitting Technique for a Second Life Avatar

Dohi

Ishizuka

2011

Lecture Notes in Computer Science

View full text Add to dashboard Cite

show abstract

“…For example, the Grassmannian manifold of subspaces of a vector space has been used in face recognition [5] and the Kendall manifold of shapes has been used to model face shape [6].…”

Section: Introductionmentioning

confidence: 99%

Manifold-based constraints for operations in face space

Patel

Smith

2016

Pattern Recognition

View full text Add to dashboard Cite

In this paper, we constrain faces to points on a manifold within the parameter space of a linear statistical model. The manifold is the subspace of faces which have maximally likely distinctiveness and different points correspond to unique identities. We provide a detailed empirical validation for the chosen manifold.We show how the Log and Exponential maps for a hyperspherical manifold can be used to replace linear operations such as warping and averaging with operations on this manifold. Finally, we use the manifold to develop a new method for fitting a statistical face shape model to data, which is both robust (avoids overfitting) and overcomes model dominance (is not susceptible to local minima close to the mean face). We provide experimental results for fitting a dense 3D morphable face model to data using two different objective functions (one underconstrained and one with many local minima). Our method outperforms generic nonlinear optimisers based on the BFGS Quasi-Newton method and the Levenberg-Marquardt algorithm when fitting using the Basel Face Model.

show abstract

“…Some of the primary benefits of such models include potentially improved robustness to pose and illumination changes during recognition [3], estimation of 3D facial shape from 2D images [2,20], and motion capture [13]. Given this emerging popularity, a great need exists for rigorous and standardized 3D dynamic facial data sets that the computer vision community can use for experimentation.…”

Section: Introductionmentioning

confidence: 99%