SeeVoice Non Real-Time: How It Works

SeeVoice Non Real-Time converts recorded voice into synthetic video clips with animated avatars (talking characters).
The product is based on SeeStorm Phoneme Recognition / LipSync technology, and is delivered as a server software product.

SeeVoice, installed on servers of Service Provider, processes the incoming voice fragments to extract phonemes from the speech and to convert them into commands for animation of an indicated 3D character.
Administrative server performs all content processing management.
The resulting synthetic video with the avatar speaking the message is ready to be sent to recipient's mobile handset or to be used in another way defined by the given service structure.

The input is a voice file in AMR/PCM/G.711 format, and the IDs of the chosen avatar, background picture and emotional state.
The output is a 3GP (H.263+AMR) video file.
Other output formats are can be provided on your request.

The length of incoming voice messages is quite flexible, beginning from 1 sec.
The size of resulting message is adjustable: it depends on the desired video quality, resolution, frames per second rate and the incoming voice message length.

SeeVoice can be integrated with some Text-to-Speech (TTS) software to enable conversion of text into synthetic video. It makes possible mobile services where users send common SMS - which are received as MMS with talking characters.