From 2fe4d3c3a0bf00e51d3d89566c812bedc95ce281 Mon Sep 17 00:00:00 2001 From: Victor Kuznetsov Date: Tue, 1 Sep 2026 15:33:26 -0700 Subject: [PATCH] Correct stale claims in the user-facing docs README omitted doubao from both places it lists supported video marks; the mark has been registered since 2026-08-28 and docs/cli.md and supported-signals.md already carry it. README also attributed the 2026-07-29 two-clip calibration to video-synthid-oracle.csv, which holds two rows from one carrier dated 2026-07-31; that calibration was never recorded in a tracked manifest, so say so rather than implying it is reproducible. "Both surviving profiles" predates chroma-zimage; there are three. "The SDXL profiles were removed" reads as false next to the shipped sdxl-zimage -- docs/synthid.md already had the right wording, naming the standalone SDXL and ControlNet profiles, so use it in the other two places. The Hub card declared a datasets entry for hugging-face/open-images, which does not exist and rendered a dead link on the model page. Open Images stays named in the prose, where it is a source rather than a Hub id. Co-Authored-By: Claude Opus 5 --- README.md | 13 ++++++++----- docs/known-limitations.md | 5 +++-- docs/photo-classify-hf/README.md | 1 - docs/synthid.md | 2 +- 4 files changed, 12 insertions(+), 9 deletions(-) diff --git a/README.md b/README.md index ce9cde9..4f1133a 100644 --- a/README.md +++ b/README.md @@ -7,8 +7,8 @@ Remove AI provenance marks from images and video you generated yourself: - C2PA, EXIF, XMP, IPTC, and related AI metadata. Video support covers provenance identification, complete visible-plus-metadata -cleaning, directory batches, visible Sora, Veo, Seedance, Dola, Hailuo AI, and -Kling AI mark removal, and oracle-certified VAE regeneration for video SynthID +cleaning, directory batches, visible Sora, Veo, Seedance, Doubao, Dola, +Hailuo AI, and Kling AI mark removal, and oracle-certified VAE regeneration for video SynthID removal. > [raiw.cc](https://raiw.cc) runs this library as a hosted service, with the GPU @@ -446,7 +446,7 @@ invisible removal. detail. - Visible video removal recognizes the moving Sora 2 wordmark, the current Veo diamond plus legacy `Veo` text, the Seedance boxed `AI` label, and the fixed - Dola, Hailuo AI, and Kling AI labels. It does not recognize the older Sora Turbo + Doubao, Dola, Hailuo AI, and Kling AI labels. It does not recognize the older Sora Turbo corner swirl or unregistered layouts from those providers. The classical OpenCV backend can smear structured backgrounds; use MI-GAN or LaMa when recovery quality matters. @@ -486,8 +486,11 @@ reinterpret the pixel result returned `UNAVAILABLE`; that follow-up was not a verifier rerun and does not invalidate the built-in verdicts. A 2026-07-31 full-clip check on a public eight-second Veo sample found `0.10` still detected and `0.15` not detected, so `0.15` is now the certified default. The -reproducible hashes and verdicts live in -`data/evaluations/video-synthid-oracle.csv`. +reproducible hashes and verdicts for that full-clip check -- one carrier, two +rows -- live in `data/evaluations/video-synthid-oracle.csv`. The earlier +two-clip calibration is narrative only: its verdicts were not recorded in a +tracked manifest, so treat the certified default as resting on the 2026-07-31 +rows. ## Documentation diff --git a/docs/known-limitations.md b/docs/known-limitations.md index 323df8b..dbd4265 100644 --- a/docs/known-limitations.md +++ b/docs/known-limitations.md @@ -274,8 +274,9 @@ per-stage schedule, CFG 1.0 and CUDA. ### Small images are processed at their native size There is no minimum-resolution floor. It existed to enlarge small inputs toward -SDXL's ~1024 training resolution and was removed with the SDXL profiles, which -never applied it anyway. Both surviving profiles run at native geometry, so a +SDXL's ~1024 training resolution and was removed with the standalone SDXL and +ControlNet profiles, which never applied it anyway. The surviving `sdxl-zimage` +is a different profile and is unaffected. All three surviving profiles run at native geometry, so a small input is neither enlarged before diffusion nor restored afterward. `--max-resolution` still caps very large inputs, and only ever scales down. diff --git a/docs/photo-classify-hf/README.md b/docs/photo-classify-hf/README.md index 2c34711..8eedc16 100644 --- a/docs/photo-classify-hf/README.md +++ b/docs/photo-classify-hf/README.md @@ -13,7 +13,6 @@ tags: - photo - provider-attribution datasets: - - hugging-face/open-images - detection-datasets/coco --- diff --git a/docs/synthid.md b/docs/synthid.md index ac4a904..5e1239d 100644 --- a/docs/synthid.md +++ b/docs/synthid.md @@ -1092,7 +1092,7 @@ a resolution trend here; measure it. **Current implication:** the old floor table remains evidence about the dated test set, not the current resolver. The SDXL and ControlNet profiles it measured no longer exist; the shipped defaults are defined in `watermark_profiles.py`, and -both surviving profiles run face repair as a built-in second stage rather than as +all three surviving profiles run face repair as a built-in second stage rather than as an optional restore. Removal near a threshold remains seed dependent, so reproducible verification requires a fixed seed.