Tiled Prompts: Overcoming Prompt Underspecification in Image and Video Super-Resolution
Bryan Sangwoo Kim, Jonghyun Park, Jong Chul Ye · 2026
Read Tiled Prompts: Overcoming Prompt Underspecification in Image and Video Super-Resolution by Bryan Sangwoo Kim, Jonghyun Park, Jong Chul Ye (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-03342.txt.
Description
One global caption misguides image upscaling tile by tile, and Bryan Sangwoo Kim, with 2 co-authors, fixes that mismatch for super-detail. Text-conditioned diffusion models use prompts as meaning priors, while pipelines tile the latent image to reach high resolutions. A single coarse caption misses localized details and supplies locally irrelevant guidance, producing substandard tiles, hallucinations, and visible artifacts in the enlarged output compared with what local description could support.
Their Tiled Prompts system makes a prompt given to each latent tile. Super-detail then runs under locally text-conditioned posteriors, resolving both omissions and irrelevant guidance with minimal overhead. Tests on high-detail real-world images and videos showed consistent gains in perceptual quality and fidelity. Hallucinations and tile-level artifacts fell relative to global-prompt earlier methods. A project page accompanies the work.
Catalog record
| CREATOR | Bryan Sangwoo Kim, Jonghyun Park, Jong Chul Ye |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 41 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-03342