A novel set of tools developed by a team of EU-funded researchers has given artists and media professionals the means to create a quality audio-visual animation at a fraction of the time and cost. The tools allow a user to choose their character's facial expressions, physical attributes, and even the pitch of their voice. The 'intelligent content' technology was developed under the SALERO ('Semantic audio-visual entertainment reusable objects') project, which was funded EUR 8.85 million through the EU's 'Information society technologies' Theme of the Sixth Framework Programme (FP6).
Headed by Joanneum Research in Austria, the 13-partner consortium has successfully developed a total of 24 applications, tools and showcases that enable a user to create, manage, edit, retrieve and deliver content for their own computer game or animation. This means being able to create characters, objects, sounds, language sets and behaviours in a context of your own choosing and with little technical know-how.
The objective for SALERO's researchers, artists and creative media professionals was to make computer gaming and animation more accessible by speeding up the production process, and making it easier and less expensive. This would allow a range of simulations, such as training content for healthcare professionals, to be produced more efficiently, and to allow these simulations to adapted to different scenes or entirely different projects.
Launched in 2006 and finalised in late 2009, the team succeeded in creating a refined, self-adaptive and semi-automated product that is already attracting considerable attention. Joanneum Research's Dr Georg Thallinger noted that studios are particularly excited about the audio and animating tools. After all, these tools are likely to save industry professionals time and money, in the kind of way that the phenomenon of stock photo libraries have made images much more accessible.
'It means you do not need as many voice actors, which is expensive, and it does not take as long to create a wide variety of characters expressing a broad range of emotions,' said Dr Thallinger.
By using the synthetic emotional model, for example, a user has access to the 'activation' (strength of the emotion) and 'evaluation' (balance control) variables that allows them to animate many different expressions. What's more, by using the Maskle tool, these same emotions can easily be transferred to another character's face.
A user can also choose the gender, ethnicity, age and weight of their character. These variables also influence the character's gait and body movement, which are determined automatically depending on the type of surface the character is on.
'The animator can then simply point to an area where the character must go, and the software chooses the path and animates the character variables, and the quality of the surface along the route,' explained Dr Thallinger.
Thanks to the audio processing tools, the technology can transform a voice recording to reflect a different gender, age, speed, timbre, pitch, and even emotional stress (e.g. from happy to sad). If you do not have a voice recording, a computer-generated one can be created for you.
The team has also established a system where a character's emotional stress is synchronised with tone of voice. 'It leads to a much more realistic and convincing animation when movement is synched across voice and video in this way,' added Dr Thallinger.
Information on the various tools is available on the SALERO website including several animated productions. The SALERO project brought together researchers and industry actors from Germany, Ireland, Spain, Austria, Finland and the UK.