Navigating the manifold: A transformative approach to AI image generation
Using an external starting image and sparse action prompts, Transformative AI treats image generation as iterative navigation through an irregular, high-dimensional manifold. This user-guided process preserves expressive development and explores regions apparently inaccessible to text-to-image workflows. I illustrate this with an example and compare it to a text-to-image equivalent.
Authors
- Alec D. MacCall (ORCID: https://orcid.org/0000-0002-0805-7787)
Institutions
- National Oceanic and Atmospheric Administration (US)
- NOAA Oceanic and Atmospheric Research (US)
- Farallon Institute (US)
Publication Details
- Journal
- Patterns
- Published
- 2026-04-01
- DOI
- https://doi.org/10.1016/j.patter.2026.101539
- Primary Topic
- Generative Adversarial Networks and Image Synthesis
- Type
- article
- Field-Weighted Citation Impact
- 0.00