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 the oracle-measured Google strength floor to qwen-zimage

Browse Source 
The resolution curve's 0.154 top left the 4.33 MP CJK-sign fixture
SynthID-detected x3 in Gemini on the full production path (visible
stage -> qwen-zimage seed 0 -> resize-back -> metadata strip), with a
valid pixel-identical stripped control in the same session
(2026-08-18). Google-provenance content now resolves to the flat
QWEN_ZIMAGE_GOOGLE_STRENGTH 0.30 floor instead of the area curve -
anchors at 0.30 measured clean in Gemini on two fixtures (CJK sign +
18-face) at 3/3 checks across two work accounts, and stayed clean
under the vae-glyphs donor layer. openai/unknown content keeps the
curve; an explicit strength still wins.
This commit is contained in:
Victor Kuznetsov
2026-08-18 22:34:41 -07:00
parent 2298282b8b
commit 3cd1e47935
8 changed files with 56 additions and 15 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/known-limitations.md
+6 -2
View File
@@ -178,8 +178,12 @@ The two profiles resolve an unset strength differently, because different things
were measured for each.
`qwen-zimage` reads it from image area, through the resolution-adaptive denoise
curve. The vendor is deliberately ignored: the curve, not the issuer, is what was
calibrated.
curve. The vendor is deliberately ignored for openai/unknown content: the curve,
not the issuer, is what was calibrated. One measured exception (0.27.2):
Google-provenance content takes the flat `QWEN_ZIMAGE_GOOGLE_STRENGTH` 0.30 floor
- the curve's 0.154 top left a 4.33 MP fixture oracle-detected on the full
production path (2026-08-18), while 0.30 anchors measured clean on two fixtures
in two accounts.
`sdxl-zimage` reads it from the C2PA issuer, on a flat ladder:
docs/synthid.md
+12
View File
@@ -769,3 +769,15 @@ reproducible verification requires a fixed seed.
8. Jiang et al. (2025). **VideoMarkBench: Benchmarking Robustness of Video
Watermarking.** arXiv:2505.21620.
https://arxiv.org/abs/2505.21620
**Google floor on qwen-zimage (0.27.2, 2026-08-19).** The bracket above did not
survive the full production path: on the 4.33 MP CJK-sign fixture (visible-stage
sparkle removal -> qwen-zimage seed 0 at the curve's 0.154 top -> resize-back ->
metadata strip), the Gemini verifier detected SynthID x3 with a valid
pixel-identical stripped control in the same session. Google-provenance content
now takes a flat `QWEN_ZIMAGE_GOOGLE_STRENGTH = 0.30` floor instead of the area
curve - 0.30 anchors measured clean in Gemini on two fixtures (CJK sign +
18-face) at 3/3 checks across two work accounts, and stayed clean with the
vae-glyphs donor layer on top; openai/unknown content keeps the resolution
curve. An explicit strength still wins. Certification artifacts: raiw-app
`data/certification/text-restoration-2026-08-18/` (the failing re-baseline).