CalvinBackup/remove-ai-watermarks
1
0
Fork 0
mirror of https://github.com/wiltodelta/remove-ai-watermarks.git synced 2026-08-19 20:17:12 +02:00
Code Issues Packages Projects Releases Wiki Activity

Add verified text restoration

Browse Source
This commit is contained in:
Victor Kuznetsov
2026-08-15 12:25:35 -07:00
parent 8c00525946
commit 432b63b6d7
21 changed files with 974 additions and 255 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/known-limitations.md
+11 -10
View File
@@ -69,15 +69,15 @@ difficult faces. The measurements and their OCR and oracle caveats are tracked
in [`data/evaluations/fidelity/`](../data/evaluations/fidelity/README.md).
A global Z-Image Turbo prototype preserved text substantially better at low
strength, but it has no useful cross-provider operating point and is not a
supported profile. The evaluated text restorers also remain research-only:
fresh-font and silhouette variants visibly changed typography, while the
higher-fidelity `vae-glyphs` route still requires verified strings, line
geometry, a separately generated donor, and an independently clean global
anchor. Automatic OCR and line-box proposals are not reliable enough to remove
those requirements, and the exact oracle results do not establish a general
mask, seed, or provider operating range. Qwen-Image-2.0 is hosted-only and
exposes no equivalent low-strength denoise control. Exact experiments, controls,
and pass rates are kept in
supported profile. Automatic text restorers also remain research-only:
fresh-font and silhouette variants visibly changed typography. The higher-fidelity
`vae-glyphs` route is available only as an experimental opt-in with verified strings
and line geometry. It builds its donor internally but still requires an independently
clean global anchor. Automatic OCR and line-box proposals are not reliable enough to
remove those requirements, and exact oracle results do not establish a general mask,
seed, runtime, or provider operating range. Qwen-Image-2.0 is hosted-only and exposes
no equivalent low-strength denoise control. Exact experiments, controls, and pass
rates are kept in
[`text-protection-research.md`](text-protection-research.md) and the
[`fidelity` evaluation record](../data/evaluations/fidelity/README.md).
@@ -192,7 +192,8 @@ certified at a fixed seed. The live resolver is
| `qwen-zimage` | CUDA only, large model stack, and limited broad certification across seeds and content. |
| `sdxl-zimage` | CUDA only. Its strength ladder is flat per vendor, not a resolution curve, because flat values are what was measured. |
The evaluated text-restoration prototypes are not optional production stages.
Only manually verified `vae-glyphs` is an optional production stage, and it is
experimental rather than a default.
OCR plus LaMa recovered literal poster text but changed fonts and worsened whole-image
fidelity. Restricting it to OCR-mismatched lines improved the tradeoff but still
left a local shadow on one poster. The published AnyText2 SD1.5 checkpoint