OVERVIEW
A 3D audio-visual renderer has been developed to implement Momentum's
patented face animation technology for mobile and videophone applications.
The renderer allows for high-quality, speech-synchronized lip and head animation with minimum latency.
Real-Time Avatar for Video Telephony
Momentum offers a novel technology to provide a real-time avatar-based video call
solution for 3G mobile and videophone service providers. During an avatar video
call, the user's voice will be processed in real-time in order to drive Momentum's
facial animation rendering engine on the mobile or videophone service providers servers.
The lip-synchronized and photo-realistic live avatar video stream will be transmitted
over the network to the other user's phone.
Features
-
Auto lip-synchronization, facial expression and head motion
-
Standard set of avatars as well as automatic personalized avatar generation
|
|
Show details: This diagram shows the outline of the real-time avatar system solution.
As the speech stream arrives at the telecom service provider's system (1),
the stream is transmitted to the Avatar generation subsystem (2) for subscribers of the avatar service.
The real-time speech recognition module generates input for the real-time facial animation rendering engine (3).
The output video is compressed to a suitable format for delivery (4). The original speech stream is combined with the avatar video and finally sent
to the receiver (5).
Show demo: Here is a small demonstration of the avatar video rendering
Text-to-Speech for Avatar Rendering
Momentum also offers Text-to-Speech (TTS) technology for real-time avatar-based rendering
solution for mobile and videophone service providers. Momentum's TTS software automatically converts text input
to speech, which can be used to drive Momentum's facial animation rendering engine.
Features
-
TTS engine to create a speech sample and viseme based lip-synced render from text only
-
Rapid, automatic, low-cost video generation
-
Auto lip sync, facial expression and head motion
-
Standard set of avatars, photorealistic or cartoon
|
|
Show demo: Here is a sample output of the TTS engine:
Hide details
