Nvidia Uses Neural Network to Render 3D Environments

AlphaAtlas

AlphaAtlas

[H]ard|Gawd
Staff member
Joined
Mar 3, 2018
Messages
1,713
Nvidia claims they've developed software that can take a series of input videos, and generate a unique, but visually similar 3D representation of the world depicted in them. While Unreal Engine 4 is used somewhere in the process, the heart of the software is a neural network that can extract a "high-level descriptions of a scene" from footage and fill in the details. The company created a demo out of city driving footage, and claims it "allows attendees to navigate a virtual urban environment that is being rendered by this neural network."

Check out the rendering tech in the video here.

This AI breakthrough will allow developers and artists to create new interactive 3D virtual worlds for automotive, gaming or virtual reality by training models on videos from the real world. This will lower the cost and time it takes to create virtual worlds. The work was developed by a team of NVIDIA researchers led by Bryan Catanzaro, Vice President of Applied Deep Learning at NVIDIA... For training, the team used NVIDIA Tesla V100 GPUs on a DGX-1 with the cuDNN-accelerated PyTorch deep learning framework, and thousands of videos from the Cityscapes, and Apolloscapes datasets.
 
Ultima99 said:
Great, how long before we have fake videos of politicians murdering people in smear campaigns during elections?
Click to expand...
I was thinking the same sort of thing when I was watching the video. This tech is a pandora's box. Between this, AI facial mapping and AI voice modulation, imagine where all this tech will be in 5 or 10 years.

Scary stuff when you think about it.
 
Drop the doom and gloom. This is a remarkable step for videogames. If a dev can go out and film an entire city, or say all of yellowstone park and then just crunch it I to a usable game environment this will leapfold game production on an order of magnitude. Instead of thousands of human hours focused on world building they could use that manpower on the other aspects of a game that make it fun and rewarding. Still have to see it in practice but it is a noteworthy accomplishment at the least.
 
oldmanbal said:
Drop the doom and gloom. This is a remarkable step for videogames. If a dev can go out and film an entire city, or say all of yellowstone park and then just crunch it I to a usable game environment this will leapfold game production on an order of magnitude. Instead of thousands of human hours focused on world building they could use that manpower on the other aspects of a game that make it fun and rewarding. Still have to see it in practice but it is a noteworthy accomplishment at the least.
Click to expand...

Yeah I thought about that and it will be cool to see where games take this tech, but it's at our own peril to ignore the potential downsides.
 
Donald Bell said:
I was thinking the same sort of thing when I was watching the video. This tech is a pandora's box. Between this, AI facial mapping and AI voice modulation, imagine where all this tech will be in 5 or 10 years.

Scary stuff when you think about it.
Click to expand...

You could have almost impossible to discredit fakes, and the next worry becomes when actual guilty people claim that real video is fake.
 
Ultima99 said:
Great, how long before we have fake videos of politicians murdering people in smear campaigns during elections?
Click to expand...
You already can convince half the country a X did Y with no evidence or laughable evidence as seen in past few years. So it's a short trip.
 
  • Like
Reactions: TAP
like this
You must log in or register to reply here.
Back
Top