Q: What is NVIDIA Canvas?
NVIDIA Canvas is a painting application that uses AI to generate landscape images from simple brush strokes. This enables artists to rapidly ideate, a daunting task at the beginning of a project when faced with a blank canvas.
Q: Who is NVIDIA Canvas for?
Canvas is for artists who need to conceptualize landscapes for their projects. For example,
- Concept artists can rapidly explore new ideas
- Archviz designers can quickly draft backdrops and environments for their buildings
- Creators can rapidly paint a landscape rather than searching hours for the perfect stock photo
Moreover, Canvas is accessible so that any RTX user who wants to draw amazing landscapes, either for work or play, can do so with simple brush strokes and the power of AI.
Q: Where can I get NVIDIA Canvas?
You can download NVIDIA Canvas beta here: www.nvidia.com/canvas-app
Q: How much does NVIDIA Canvas cost?
Canvas is free to download. It requires RTX Tensor Cores to function and is supported only on NVIDIA RTX GPUs.
Q: What key features does Canvas include?
- AI-assisted painting
- 15 material brushes
- 9 styles to modify the look of the output image
- Export painting to .PNG or layered .PSD file format to import into graphic design applications that support .PSD such as Adobe Photoshop.
Q: What are the requirements to run NVIDIA Canvas?
NVIDIA GPUs: GeForce RTX, NVIDIA RTX, Quadro RTX, TITAN RTX
NVIDIA Driver: 460.89 or later, NVIDIA Studio Drivers are recommended.
Windows 10 2004 or later
- Mouse and keyboard
- Optional: touch screen
- Optional: tablet and stylus (e.g., Wacom Cintiq or Intuos Pro)
Q: Can I use my own images or photos to affect the style and look of my paintings?
This is a feature we plan to implement in a future update.
Q: I would like a new feature to be integrated into Canvas
We would love to hear your feedback about your experiences and needs to help improve future versions of Canvas. Please provide the feedback on the form: www.nvidia.com/canvas-feedback or just click the feedback button in the application:
Q: How do I provide feedback about issues I am having?
We would love to hear your feedback about your experience, so we can improve Canvas. Please fill out the feedback form or join a discussion on our Forum.
- Feedback form: http://nvidia.com/canvas-feedback
- Forum: www.nvidia.com/en-us/geforce/forums/nvidia-canvas
- Feedback button in the application
Q: How long will NVIDIA Canvas be supported?
We plan to make frequent updates to Canvas as we gather feedback from the community of artists and implement them. You can find out more about the latest or share feedback below:
Q: Where can I get help with using NVIDIA Canvas?
There is a user guide to help you get started with using Canvas here: nvidia.com/canvas-guide
You can also join the discussion with the community here: www.nvidia.com/en-us/geforce/forums/nvidia-canvas
Q: How does the AI in Canvas work?
NVIDIA Canvas uses a GAN (Generative Adversarial Network) to turn a rough painting of a segmentation map into a realistic landscape image. 5 million photographs of landscapes were used to train the network on an NVIDIA DGX.
Q: Is Canvas related to GauGAN?
Canvas is built on the same research core that NVIDIA showed in GauGAN.
Q: Does NVIDIA Canvas replace any apps I currently use in my workflow?
No. NVIDIA Canvas is designed to supplement your workflow. It’s main power is that it enables you to rapidly ideate and visualize new landscapes or environments. The result can then be exported to PNG or layered PSD so you can refine and build on the work in specialized graphic design applications, such as Adobe Photoshop.
Q: Do I need to be an AI expert to use NVIDIA Canvas?
No. You don’t need any knowledge about AI or neural networks to use NVIDIA Canvas. Canvas is designed as an artist-first application.
Q: Can I only paint landscapes? Does the app support other kinds of art creation?
Currently landscapes and environments can be painted in NVIDIA Canvas. This is because the AI in NVIDIA Canvas has been trained on millions of landscape images. We are looking at various ways to expand the functionality of Canvas. Let us know your feedback here: www.nvidia.com/canvas-feedback