Turning a timelapse into a “flowing” painting in the style of one of the great masters doesn’t necessarily require 80 hours of frame-by-frame iPad photography with the Prisma app. For artist Danil Krivoruchko, all it took was a high-tech computer algorithm.
The algorithm, developed by researchers at Cornell University, uses artificial neural networks to achieve “artistic style transfer for videos.” In this case, Krivoruchko used it to turn a timelapse captured in NYC into a moving painting in the styles of Van Gogh and Picasso. It’s a far cry from the painstaking process it took to create this Prisma timelapse.
Here’s the “source” timelapse:
And here is the transformed version:
In case you want try this yourself, Danil has actually posted the code to GitHub. But be warned: it’s going to take some serious processing power, a copy of the Ubuntu operating system, and time for some After Effects editing before it’s all said and done. This isn’t a quick project, but we would love to see some of our favorite timelapses “reskinned” this way.
from PetaPixel http://petapixel.com/2016/07/26/computer-algorithm-turns-nyc-timelapse-moving-van-gogh-painting/
No comments:
Post a Comment