Machine learning generally is a fantastic tool for creators; however, integrating AI into your workflow is a problem for those who can’t code. A brand new program referred to as Runway ML goals to make this process simpler by offering artists, designers, filmmakers, and others with an “app store” of machine learning applications that may be activated with a couple of clicks.
Say you are an animator on a budget who desires to turn a video of a human actor right into a 3D model. Instead of hiring expensive motion-capture tools, you might use Runway to apply a neural network referred to as “PosetNet” to your footage, creating wireframe models of your actor that may then be exported for animation.
Or say you want to remove a coffee cup that was accidentally left in a shot in your high-budget fantasy TV drama. You possibly can edit it out the traditional way, painting over the cup by hand, or you would run your footage through a machine learning segmentation model, which might automatically highlight different objects in each frame to make your job easier.
Examples like these are just the tip of the iceberg for Runway, which co-founder Cristóbal Valenzuela describes as a radically egalitarian instrument. “Machine learning is very exclusive technology,” Valenzuela tells The Verge. “However, I need to make issues more inclusive; to get folks from different backgrounds sitting around the table and utilizing these models.”