If you’ve ever wanted to dig into the latest in deep learning research, now’s your chance.
NVIDIA has launched AI Playground, an online space where anyone can experience our research demos firsthand.
“Research papers have new ideas in them and are really cool, but they’re directed at specialized audiences — we’re trying to make our research more accessible,” said Bryan Catanzaro, vice president of applied deep learning research at NVIDIA. “The AI Playground allows everyone to interact with our research and have fun with it.”
The Image InPainting model, presents an AI imaging technique that edits out unwanted content and replaces the resulting holes with realistic computer-generated alternatives. See the results yourself by uploading and editing your own photos to the Image InPainting demo. This work was first published last fall at the European Conference on Computer Vision in Munich, Germany.
The Artistic Style Transfer demo takes the aesthetic of one image and applies it to another, with realistic results. For example, a user can upload a photo of their friends and convert it to look like it was painted in the style of van Gogh style. This research was introduced in August at SIGGRAPH, in Vancouver.
The Photorealistic Image Synthesis model is a new deep learning-based system that uses conditional generative adversarial networks to generate photorealistic images and virtual environments. The demo is based on research presented in June at the Conference on Computer Vision and Pattern Recognition, in Salt Lake City.
More demos will be added in the months ahead.
According to Catanzaro, AI Playground is for anyone with interests that span from photo editing to deep learning.
“NVIDIA is uniquely positioned between visual computing and AI,” said Catanzaro. “That nexus is reflected in our research, and in the demos you’ll see on the AI Playground.”
Sign up to receive updates on NVIDIA Research and future AI Playground demos.