Watch these Lips: Motion Segmentation of Human Speech and Voice Organs.
Sven FriedlWalter ZinkBernhard FröbaThomas WittenbergPublished in: VMV (2002)
Keyphrases
- motion segmentation
- image sequences
- model selection
- moving objects
- motion estimation
- multibody
- missing data
- text to speech
- optical flow
- epipolar geometry estimation
- dynamic scenes
- motion model
- flow field
- speech recognition
- spectral clustering
- speech signal
- feature tracking
- automatic speech recognition
- data mining
- fundamental frequency
- multiple moving objects
- face recognition
- video sequences
- emotion recognition
- emotional state
- structure from motion
- motion field
- vector field