Skip to content

[codex] Add spread v2 image watermark payload mode#5

Draft
CEQ151 wants to merge 2 commits intomasterfrom
codex/mlwm-v2-ux-batch-preload
Draft

[codex] Add spread v2 image watermark payload mode#5
CEQ151 wants to merge 2 commits intomasterfrom
codex/mlwm-v2-ux-batch-preload

Conversation

@CEQ151
Copy link
Copy Markdown
Owner

@CEQ151 CEQ151 commented Apr 27, 2026

Summary

  • add frequency-spread-v2 image watermarking with keyed fingerprint64 payloads by default
  • keep text16 compatibility mode and route low-visibility profiles through the new DWT/DCT spread codec
  • surface payload mode, fingerprint, confidence, BER, and codec diagnostics through helper, IPC, and UI
  • add spread-v2 unit tests and a local benchmark runner for clean/JPEG/resize/blur/crop/FPR checks

Validation

  • ..venv-ml\Scripts\python -m unittest discover -s blind_watermark\tests
  • npm run build
  • npx eslint --quiet src/main/index.ts src/preload/index.ts src/preload/index.d.ts src/renderer/src/components/ImageWatermarkPanel.tsx src/renderer/src/i18n.ts

Notes

  • Draft PR because resources/bin/bwm_helper.exe has not been rebuilt yet; packaged production builds still need a helper rebuild before release.
  • Local smoke on 5 train images: clean/JPEG85 20/20, original FPR 0/20, PSNR >= 51.331 dB, SSIM >= 0.997982.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant