← ArchivePaper2014
FaceWarehouse: A 3D Facial Expression Database for Visual Computing
Large-scale 3D facial expression database with bilinear model fitting enabling data-driven face tracking, synthesis, and editing.
Abstract
We present FaceWarehouse, a database of 3D facial expressions for visual computing applications. We use Kinect, an off-the-shelf RGBD camera, to capture 150 individuals aged 7-80 from various ethnic backgrounds. For each person, we captured the RGBD data of her different expressions, including the neutral expression and 19 other expressions such as mouth-opening, smile, kiss, etc. For every RGBD raw data record, a set of facial feature points on the color image such as eye corners, mouth contour, and the nose tip are automatically localized, and manually adjusted if better accuracy is required. We then deform a template facial mesh to fit the depth data as closely as possible while matching the feature points on the color image to their corresponding points on the mesh. Starting from these fitted face meshes, we construct a set of individual-specific expression blendshapes for each person. These meshes with consistent topology are assembled as a rank-3 tensor to build a bilinear face model with two attributes: identity and expression. Compared with previous 3D facial databases, for every person in our database, there is a much richer matching collection of expressions, enabling depiction of most human facial actions.
How to read this
- Category
- Dataset / 3D facial database with a bilinear model
- Contributions
- A database of 3D facial expressions for 150 individuals (ages 7-80, varied ethnicities) captured with a Kinect RGBD camera
- For each person, a neutral plus 19 expression captures, with facial feature points auto-localized and manually adjusted
- A template mesh fitted to depth and color features, yielding individual-specific blendshapes assembled into a rank-3 tensor (identity x expression) bilinear face model
- Context
- Builds on the morphable-model tradition for faces (Blanz and Vetter's 'A Morphable Model for the Synthesis of 3D Faces'), extending it to a consumer-depth-camera-captured, expression-rich bilinear formulation.Builds on: A Morphable Model for the Synthesis of 3D Faces
- Correctness
- Validated as an asset for data-driven tracking, synthesis, and editing; note the data comes from commodity Kinect RGBD with partly manual feature correction, so reconstruction fidelity is bounded by that capture setup and the chosen 19-expression set.
- Clarity
- Accessible; a first pass conveys what the database offers, a second pass pays off for the mesh-fitting and tensor (bilinear) construction details.
- How to read it
- Skim the abstract and figures first to judge fit as a resource; if you plan to use the model, do a second pass on the fitting pipeline and the rank-3 tensor decomposition, and check the expression list against your needs.
Builds on
Related work
- Animating Facial Expressions 1981 / SIGGRAPH
- A Muscle Model for Animating Three-Dimensional Facial Expression 1987 / SIGGRAPH
- 3D Shape Regression for Real-Time Facial Animation 2013 / SIGGRAPH
- Facial Performance Enhancement Using Dynamic Shape Space Analysis 2014 / SIGGRAPH
Keywords
This page summarises the entry and links to its original source. The archive never hosts or redistributes the publication itself.Show it in the full archive list →