Read more.The trained convolutional network can already operate at interactive frame rates.
Read more.The trained convolutional network can already operate at interactive frame rates.
Does it look more real? Definitely. Does it look better? Not really.
I can see this perhaps being very useful in a different game. But people play games for colourful, exciting stimulation. This just reminds me of how boring city life is.
Now, can we take a moment to appreciate how good the person playing GTA was trying to be? He clearly wanted to go on the wrong side of the road and flatten some motorcyclists....
Looking realistic VS looking good are far from what the average Joe wants / feel.
For instance for me, i have yet to see really good looking footage from any action camera, it is also the same with action cameras, here it is often really bad.
Another place you often see it is in commercial ( i want to say photography but really i should say paint-shopping )
The truth !!!!!! you cant handle the truth.
I can see how this would look better if the source data was taken on a sunny day in LA, or even better across several different weather conditions.
It would be interesting to see how this tech would translate to faces, although perhaps dystopian. And if it became common place in games set in the current day or near present/past it would inherently cause games set in the past or future to have a different art style.
Maybe it could be used by a low budget filmmaker to create scenes in a game engine and then post-process them to match the look of the rest of their footage.
That side-effect of applying the lighting and colour of a different place entirely could be a nice creative tool as well.
Thats the exact same colour values used for GTA 4, if I am not mistaken.
There are currently 1 users browsing this thread. (0 members and 1 guests)