post building the project

July 18th, 2008

Filed under: CHAMELEON PROJECT — Tina @ 6:20 pm

I left Australia in January, with my family, to embark on the ANAT synapse residency, where I was to work with my collaborators to build the beginnings of the Chameleon Project, a two year project to be finalized in early 2010. I traveled to MITMedia Lab twice to build the mind reading technology and work with Ros Picard and Rana El kaliouby of the Affective computing Group. I then travelled to the Banff Center to shoot the beginning video of the project. We then moved to London to work with Chris Frith, a neuroscientist at the Wellcome Department of Neuroimaging, as well as Hugo Critchley, chair of Psychiatry at Brighton and Susses Medical School.

To understand “Chameleon’s” complexities, we are building the work in small stages. There is many reasons for this – one – to make sure each element of the project is well researched. To provide a ‘visual language that all collaborators can understand, to provide a more modular, ‘jumping off’ point so each research may be able to take the project and use it with in their own research and experiments. To provide a platform where the work can be exhibited in progress, and audience evaluation and feedback can be gathered. 

We are now working on experiment 5 and 6.

Experiment one: creating an ‘ethnographic style short film’ to understand the most potent actions that cause emotional contagion (shot at BMNI) (still in edit)

Experiment two: creating a live emotional contagion tool that transposes live social interactions and slows it down to reveal micro-expressions. (showed at the Dana Centre, Science Museum London)

Experiment three: implementing Frith’s emotional algorithm to drive two still portraits that emotionally respond to each other. Live feed to the web (chat rooms) – to transpose text written by a man writing about the emotion they are expressing, to transpose text written by a woman writing about the emotion they are expressing. (below potential mock up for gallery space) – showed in Banff

 Experiment four: Understanding how emotions propagate in social groups


Experiment five: Building a video engine that integrates the emotional algorithm hypothesized by Chris Frith. Using higher resolution video, two people converse with each other.


Experiment six: Building a video engine that integrates the emotional algorithm hypothesized by Chris Frith. Using higher resolution video, twelve people converse with each other, infecting each other with their emotional state.

Experiment seven: Integrating the mind reading technology developed the the MIT MediaLab.

Experiment eight: exploring different displays with a more sculptural aesthetic.

experiment nine: using three networked computers and cameras to drive three emotional portraits

Experiment ten:  final work: up to five participants enter a space, triggering video portraits that aim implicate you into an emotional drama.


post trouble shooting with experiment 05

July 18th, 2008

Filed under: CHAMELEON PROJECT — Tina @ 5:23 pm

experiment 5 is a ‘duel player’. The patch plays two HD videos on two screens, and they interact with each other via emotional algorithms. I am looking at the footage of susan and kevin. I have done all the edit points for each emotion – five edit loops for each emotional expression. A the end of playing the footage, it looks ot the emotional response of the other which triggers the next bit of footage.

I figure this experiment will let us know more about timings – about when to override a bit of footage and respond to the emotional expression.

So far, its not working. I am using the emotional algorithms, but they are stuck in neutral, – also   there is no print out of what is going on, as we had in the 12 player one, so its not as easy to track what is happening. Interestingly, for the last 40 minutes, they have been stuck in neutral. Looking at the loop points, I don’t see any movement – flipping between different neutral selections, but it does seem to be playing the one loop point again and again. Christian has set off to the airport so i` am unsure what to do about it. I will look at the emotional algorithm and see whats happening.

Using ‘ISHOWU’ application, I am sending the quicktimes of the patch working and the video being triggered so he can try to work it out.

1. it plays the same emotion over and over again, rather than selecting from the five options.
2. It never gets past neutral unless I hit the randomise button.
3. One I hit the randomise button it still continues to play the same emotions over and over again.
4. I changed the emotional algorithms to test it, and make it get out of neutral to: 

post working on experiment 06 to be exhibited at UCLH.

July 18th, 2008

Filed under: CHAMELEON PROJECT — Tina @ 5:16 pm

I have been working with christian topfner, a programmer based in london to create the beginning video engine for Chameleon.

The video engine is ‘driven’ by the algorithms hypothesized by Chris Frith. I have had to create these over meeting with Chris.

In the end – the work will be networked running up to about ten computers.  We have an exhibition opening at UCLH, on July 28th. It seems to be a great system – I have had a few problem with it – but I can see where its going, and its a great tool that we can changed the emotional algorithms externally – making it more complex as we go on – and also, the edit points can be changed easily – its important that none of this is hard coded, as we need lots of room to experiment. I have spent the last week looking at edit points.

We made an application of experiment 06 on my machine. Its working – looking good – playing on my machine at 21 frames per second with sound. Will now change the files to work with out sound and see the frame rate. But as soon as I put it on the install computer, not working. I have left it for Christian to install. Works on my computer – duel 2.6 gig macbookpro, taking up a about 110 of processing power.

   Footer Anatlogo Arts Sa
Australian Government The Visual Arts Strategy