Computer Vision / Video Analytics

NVIDIA Research at CVPR 2020

Researchers, developers, and engineers from all over the world are gathering virtually this year for the 2020 Conference on Computer Vision and Pattern Recognition (CVPR). NVIDIA Research will present its research through oral presentations, posters, and interactive Q&As. 

NVIDIA’s accepted papers at this year’s online CVPR feature a range of groundbreaking research in the field of computer vision. From simulating dynamic gaming environments to powering neural architecture search for medical imaging. Explore the work NVIDIA is bringing to the CVPR community.

You can find the full schedule of NVIDIA’s activities here.

Here are a few featured papers & projects:

Synthesizing High-Resolution Images with GANs

Developed by NVIDIA researchers, StyleGAN2 yields state-of-the-art results in data-driven unconditional generative image modeling.

Authors: Tero Karras, Samuli Laine, Miika Aittala, Janne Hellsten, Jaakko Lehtinen, and Timo Aila

Paper  | CVPR Talk | Blog

Learning to Simulate Dynamic Environments With GameGAN

Trained on 50,000 episodes of the game, a powerful new AI model created by NVIDIA Research, called NVIDIA GameGAN, can generate a fully functional version of PAC-MAN — without an underlying game engine. That means that even without understanding a game’s fundamental rules, AI can recreate the game with convincing results.

GameGAN is the first neural network model that mimics a computer game engine by harnessing generative adversarial networks, or GANs. Made up of two competing neural networks, a generator and a discriminator, GAN-based models learn to create new content that’s convincing enough to pass for the original.

Authors: Seung Wook Kim. Yuhao Zhou, Jonah Philion, Antonio Torralba, and Sanja Fidler 

Paper  | CVPR Talk | Blog

See a list of all of this year’s papers, workshop & tutorials, featured projects, and job opportunities at NVIDIA, on this year’s, NVIDIA AT CVPR 2020, site.

Discuss (0)