Method and apparatus for cross-modal predictive coding for talking head sequences
US5907351A · kind A · utility
Assignee
Inventors
Key dates
| Filing date | Oct 24, 1995 |
| Grant date | May 25, 1999 |
| Priority date | — |
| Expiry date | Oct 24, 2015 |
Classification
- Technology area (CPC H)Electricity
- CPC primaryH04N19/20
- WIPO fieldAudio-visual technology
- WIPO sectorElectrical engineering
Abstract
A method and apparatus for transmitting and remotely displaying the audio and visual portion of a person speaking so that the audio and visual signals are synchronized. The audio signal is constantly transmitted to the receiver and is also used to create a predicted image of the lips of the talking head. The actual lip image is compared to the predicted lip image. Based upon this comparison, it is determined which of three signals is to be transmitted to the receiver: no signal corresponding to the video signal, a signal corresponding only to the differences between the actual lip image and a predicted lip image, or the actual lip image. The receiver reconstructs a lip image based upon the audio signal received and the signal received, if any, corresponding to the video image and inserts it into the previously received video frame or modifies the previous frame accordingly.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.