I recall a post on Twitter by Scott Robertson talking about NVIDIA Canvas a while back… Scott is well known as a Concept Designer, Author/Educator, Artist, and is an early adopter of new tools and solutions that can aid the artistic process. The product page for Canvas states:
NVIDIA Canvas: Use AI to turn simple brushstrokes into realistic landscape images. Create backgrounds quickly, or speed up your concept exploration so you can spend more time visualizing ideas.
There are various lenses and motivations to consider when judging solutions such as Canvas, and it’s useful to consider the following user stories:
- As a film/tv director I would like to generate ideas for a location scout
- As a creative director at an ad agency, I would like to block out settings for a product shoot
- As a landscape concept artist, I would like to create a variety of mock-ups for art director review
- As an art director, I would like to generate rough concept art for a 3D artist to author an environment
Having worked across several films, animations, and video games, I cannot overstate the need for solutions that aid in the ideation process. Regardless of medium, we tend to look for reference photos, and video footage, to help us nail down a target that can support a broader narrative or design. Using a solution like Canvas, can help us explore a variety of ideas with unprecedented ease, which is a practical need in professional circumstances.
While some of us could debate that Canvas is not quite generating results indistinguishable from a real photograph, it can be vastly more useful than a rough sketch. And here are some explorations by Scott Robertson in ArtStation.