The GAN-based AI technology Has Been presented in the NeurIPS conference in Montreal this week.
Tons of people around the world love video games. This sort of entertainment has taken over a great deal of houses and is the last time of a good deal of people.
As one might expect, creating a virtual 3D world is a laborious process that requires not only time and effort but also a high level of skill.
Look at any video game, and the graphics will have required tens of thousands and thousands of hours of work on the part of developers specialized in 3D rendering, lighting, textures and more.
That level of work to make virtual environments might not be necessary for much longer, but as graphics processing unit company Nvidia has developed a strong AI algorithm capable of generating photorealistic details for a virtual world.
The AI tool could not just affect the video game development industry, but may also have important applications in creating environments for virtual reality as well as for coaching self-driving cars and robots.
Bryan Catanzaro, the vice president of applied profound learning at Nvidia, explained:”We can produce new sketches that have never been seen before and render those. We’re actually teaching the model how to draw based on real video.”
If the AI tool sounds impressive, that’s because it is. Nvidia’s team has utilized a traditional machine-learning approach to identify various objects appearing in a movie scene, like cars, trees, houses, etc..
From there, it’s leveraged a more sophisticated generative adversarial network (GAN) to train a computer to generate photorealistic 3D vision to fill the outlines of objects.
In other words, users can enter the outline of a scene (indicating where different objects should be), and the algorithm will fill in the outlines with realistic information.
At this point, the images generated by the algorithm are not ideal –as some things can appear lightly warped, almost dreamlike–but the outcome is already highly advanced.
“Classical computer images render by establishing the way light interacts with objects,” added Catanzaro.
“We wondered what we could do with artificial intelligence to change the rendering process.”
The revolutionary approach was introduced at NeurIPS, an AI conference hosted in Montreal this week and has gained considerable attention already, largely from the gambling industry, which may radically improve (in terms of ease) game design.
“NVIDIA has been inventing new ways to create interactive images for 25 years, and this is the first time we can do this with a neural network,” stated Catanzaro.
“Neural networks — especially generative models — will alter how graphics are created. This will enable developers to create new scenes in a fraction of the traditional price.” Invented by Google, GANs have only existed for a few years but are already demonstrating their ability to produce realistic vision and sounds. You might remember a viral video of Trump from May 2018 calling to withdraw from the Paris climate agreement.
The video–though obviously fake–was created using GANs and demonstrated that the potentially dangerous ramifications of the technology on producing realistic and (actually fake) fake news.
Still, from a technology point of view, Nvidia’s algorithm is mighty impressive and we’re eager to see what its first real world (virtual world?) applications are.
Tons of people all over the world enjoy video games. Every day people buy new games, watch movies and videos online about games and read about video games. This kind of entertainment has taken over a lot of homes and is the past time of a lot of people.