Hi Guys,
With my supervisor permission, I am allowed to discuss now regarding our new experiment done on generating landscape designs with less user Inputs.
Less user Inputs
?
YES, With less than 5 mouse clicks for every 2 mins, could bring what user is expecting to design (imaginery scenery) almost closer to 75%.
In this process, user need NOT to know how to modify, adjust and attach parameters at any point during the process. Everything is adjusted by it self.
How
?
Using IGA, Interactive Genetic Algorithm.
Our application reads user mind and communicates with Terragen engine with user specifications to render what he is looking for. Selected twice on IEEE Evolutionary Computation.
Any Example?
http://www.youtube.com/watch?v=OISauX-XAVYhttp://www.youtube.com/watch?v=qDey8uyzvqgAny problems
?
Yes, Rendering what I can. I need to render 16 images at once which is taking almost 2 to 3 mins for every generation. Planned to implement network rendering but ran out of time now. May be in our next update.
UI interface is attached with this post. I am happy to answer any type of doubts that you guys might have. It sounds strange alright when I say this process really works but it does
.
Many thanks to Oshyan, Matt, rArcher and manymore those who share their ideas and helped me around during my research.