Living Actor™ Speech2Video is a software program that receives an audio input file containing the voice of a user and creates an output audio-video file showing an animated avatar behaving and speaking in synchronization with the input voice.
Cantoche also provides Living Actor™ Speech2Video real-time. Speech2Video real-time is a software service that receives an audio input stream containing the voice of a user and creates in real-time, an output audio-video stream showing an animated avatar behaving and speaking in synchronization with the input voice.
Main Features
- Input audio formats: Wav (audio codec: PCM, MP3, AMR), MP3; AMR, OGG
- Video formats: MP4 (H264), 3GP (H263 or H264, AMR), FLV (Sorenson Spark, MP3)
- Background pictures: PNG
Main Features: Speech2Video real-time
- Input stream: AMR, G711 (aLaw, muLaw), AAC (soon)
- Output stream: H263, H263 - 1998, H264, MPEG4 Video ES
- Background pictures: PNG
System Requirements
- OS: Linux Debian Etch, ReHat Enterprise Linux 4
- Software configuration: Python, PHP5 + mySQL + Apache
- Memory: 100MB + about 50MB / Avatar (QCIF)
- Hard-disk space: 25MB + about 20MB / Avatar (QCIF)
- Performance: Using a 2,8 GHz Intel® Pentium® 4 processor, a video file is generated in about 1 second from an audio file lasting 10 seconds.
Create a Video Avatar message with the MyMobileAvatar platform built with Living Actor™ Speech2Video:
Do you want to chat with an avatar through Skype or a web page? Ask for a demo (*).
This uses Living Actor™ Speech2Video real-time.
Skype interface © Cantoche 2009
(*) Due to significant demand, this demonstration is available only in specific cases and by request.