|
This is the webpage for Lahiru Dayananda's 3rd Year
Project at the University of York
Project Definition and Description
The aim of this project is to devise a method of building a
dataset of lip movement that can be used to improve the lip movements of a facial animation model. The mouth movement of
a speaker is to be captured and stored, along with the corresponding phonemes
that occurred with the movement. The results generated will be a dataset that
can be used to study the relationship between the position of the lips and the
phonemes being spoken.
The method by which this is to be achieved:
1)
The tracking of mouth movement. This is to be done with the use of a camera
which will be used to record clips of marked lips moving, along with the
corresponding sound.
2)
The movement of the mouth must be tracked as words/phrases/sentences are
spoken. This is made easier by the marking on the lips. Computer vision and
image segmentation algorithms can be used to track the markings on the lips. The
sound must also be extracted from the clip.
3)
The lip movement data will be recorded in some format that can represent the
displacements of the lips. The actual sounds
that correspond to the lip movements have to be broken up into phonemes. Both
sets of data must be linked together using the time as an index. The output can
be used to study the relationship between the phonemes and the lip positions.
Contents
Links and References:
- Links to online references.
- Details of offline references.
- Details of technical resources.
Documents:
- Drafts of project report chapters.
- Data and results generated.
Project Diary:
- Weekly update of work done.
- Minutes of project meetings.
Timetable:
- Timetable for project completion.
- Other commitments.
Software:
- Repository of versions of software developed.
E-mail: [email protected]
[email protected]
|