Image and model: recorded observations

2026-11-06
Eye-level pictures from the models were passed to an AI image generator with a request to keep the layout and make it look photographic. In the cafe, the image generation kept the positions and counts of doors, windows, tables and chairs, and added a ceiling, lights and materials. In the wedding venue, the number of chairs at each table changed.

2026-11-07
The path from an eye-level model picture through an AI image generator to put_image, kind: reference, was completed, returning the result as a reference image.

These are observations from these examples, not a reliability rate or a guarantee that another image will keep the layout.