Embedded Files
Decide what you want to be good at and don't ask AI to do that for you.*
Decide what you want to be good at and don't ask AI to do that for you.*
For everything else, why not?
Take almost any imgae and temporarily convert it to 3D. Then rotate the 3D object or scene to get a new view.
Generative AI is used to convert an image into 9 canonical views and these are then fused into a Gaussian Splat that the user can manipulate. The splat is then "baked" back into the Photoshop as a new layer. The new Harmonize feature uses generative AI to create realisitic lighting and shadows to blend the new view of the object seamlessly in with the rest of the scene.
This was a fun breakthrough collaboration between Adobe Research, 3D&I, and the Photoshop team.
blob
blob
This my first personal vibe-coding project where I'm creating a next-generation note-taking and todo list app that's going to be AI assisted. Stay tuned for how to access it.
NetworkFixer
NetworkFixer
I created my own bespoke AI assisted "fix everything wrong with my network app." As with all vibe-coded projects, it's a work in progress but it already helps me avoid having to do any of my own diagnoses or fixing of transient network issues
Substance 3D Viewer’s 3D model to image, powered by Adobe Firefly, gives designers more access to 3D and more control over generative AI results. This tutorial covers how to use 3D models in combination with prompts to generate awesome compositions.
I am leading the design and feature definition of the Substance 3D Stager's new generative features. These empowers users to create detailed background images from text prompts, seamlessly composite objects into scenes with intelligent perspective and lighting matching, and generate brand new 3D objects from a text prompt.
Watch what the influencers are saying!
Watch what the influencers are saying!
Check out what some of the most cutting edge 3D creators are saying about our tech!
I am helping the Adobe MINT (Machine Intelligence and New Technologies) team structure their internal explorations around generative AI + 3D. I am specifically investigating on how to best use depth maps created from assembling 3D primitives into simple scenes. This is both in the context of Adobe Firefly and future professional 3D tools.
Take a look at this video, from one of my very talented colleagues, to get a hint of how generative AI might integrate with existing 3D workflows.
This is just the beginning.
I value being good at...
I value being good at...
writing, UI design, love, human relationships, ethical decisions, art, creativity, DIY, emotional intelligence, kayaking, parenting, and bike riding. I will not rely on AI for these activities.
Page updated
Google Sites
Report abuse