Active shape models—their training and application
Computer Vision and Image Understanding
Robust Real-Time Face Detection
International Journal of Computer Vision
Automatic feature localisation with constrained local models
Pattern Recognition
Locating Facial Features and Pose Estimation Using a 3D Shape Model
ISVC '09 Proceedings of the 5th International Symposium on Advances in Visual Computing: Part I
Real-time combined 2D+3D active appearance models
CVPR'04 Proceedings of the 2004 IEEE computer society conference on Computer vision and pattern recognition
Adding facial actions into 3D model search to analyse behaviour in an unconstrained environment
ISVC'10 Proceedings of the 6th international conference on Advances in visual computing - Volume Part I
Localizing parts of faces using a consensus of exemplars
CVPR '11 Proceedings of the 2011 IEEE Conference on Computer Vision and Pattern Recognition
Hi-index | 0.00 |
We present an application which uses 3D statistical shape models to track a subject in real time using a single fixed camera. The system can handle large pose variation; variable illumination; occlusion; glasses. Since the models are 3D, the application can report pose information which may be vital in a safety context such as driving attentiveness. Two models are used in tandem, one for identity and one for facial actions, enabling the system to also estimate the user's behavioural state at a basic level. The system works directly on the captured images, with no pre-processing, and tracks the facial features using simple template matching and boundary detection. The parameters of the identity model adapt over time to the model subspace occupied by the subject, and this allows the second model to describe simple actions such as eye, brow, and mouth movement. The parameters of the actions model are then used to identify smiling, frowning, talking, and blinking using simple linear discriminants.