Extraction of 3D facial motion parameters from mirror-reflected multi-view video for audio-visual synthesis.
I-Chen LinJeng-Sheng YehMing OuhyoungPublished in: AVSP (2001)
Keyphrases
- multi view
- audio visual
- motion parameters
- visual data
- point correspondences
- image sequences
- multiple views
- multimedia
- dynamic scenes
- motion estimation
- multi modal
- least squares
- optical flow
- video sequences
- camera motion
- single view
- motion analysis
- motion model
- three dimensional
- feature points
- visual information
- d objects
- kalman filter
- facial expressions
- depth map
- video data
- moving objects
- flow field
- semi supervised
- facial images
- perspective camera
- video content
- range images
- video frames
- facial features
- motion field
- face images
- data sets
- visual features
- space time
- active appearance models
- view synthesis
- human motion
- active learning
- viewpoint
- object tracking