top of page
Virtual Image
Problem: AI-generated animations are good, but if the main subject, in this case, an Indy Formula One with particular Logos and markings, is distorted and cannot be used in real production for the client. The solution is to combine a 3D animated car composite directly in 3D software. The method of front projection allows the car to interact visually with the AI-generated 2D animation; however, this poses the problem that the AI cannot generate accurate scenes without the car present.
Solution:
A single frame is generated with a scaled version of the vehicle and used to create the AI animation. Note that the background was also generated through the Hybrid Process.
Hybrid-3d AI Generative Workflow
Google Maps Stitched low-resolution raw image
Using Google Maps multiple images were captured and stiched together in Photoshop then the final extra wide images is taken to AI upscaler to imporve the resolution and repair the mismatch and distored area.
The Google Earth image , after it is stitched in Photoshop (a combined merge of multiple screen captures),
is sent to an AI Upscaler, which enlarges and repairs it, creating a clean, high-resolution image.
bottom of page