Victor Kuznetsov df62800bf1 Add standalone dataset metadata/forensics scanner script ...

Single-file read-only scanner producing one JSONL record per image:
AI verdict via identify (metadata-only, no visible marks), full EXIF/
IPTC/XMP/PNG/JPEG/WebP/ISOBMFF inventories, C2PA store, JPEG encoder
forensics (quant tables, IJG quality, scan script, Huffman, subsampling),
XMP edit-trail fields, hashes/timestamps, macOS download provenance,
Live Photo pairing id. Runs standalone (pillow/piexif/c2pa-python) with
the library optional; supports gzip output, resume after interruption,
and head-only scanning of oversized files.

2026-07-22 17:53:02 -07:00

..

_plain_console.py

fix(scripts): drop rich import from analysis scripts (red CI after rich removal)

2026-05-31 15:41:50 -07:00

cjk_tail_probe.py

Mask faint text marks with the detector's match box, not a response threshold

2026-07-20 15:41:18 -07:00

controlnet_sweep.py

feat: controlnet pipeline for text/face-structure preservation

2026-06-03 16:59:28 -07:00

corpus_gap_scan.py

fix(scripts): drop rich import from analysis scripts (red CI after rich removal)

2026-05-31 15:41:50 -07:00

detector_response.py

Mask faint text marks with the detector's match box, not a response threshold

2026-07-20 15:41:18 -07:00

fidelity_metrics.py

fix(qwen): native-geometry img2img + pipeline-aware strength; record dropped auto/mixed/Z-Image leads

2026-06-20 21:52:56 -07:00

fill_quality.py

Mask faint text marks with the detector's match box, not a response threshold

2026-07-20 15:41:18 -07:00

invisible_quality_audit.py

chore(scripts): add invisible-removal quality audit tool

2026-06-03 19:56:49 -07:00

ladder_headroom.py

Mask faint text marks with the detector's match box, not a response threshold

2026-07-20 15:41:18 -07:00

metadata_removal_audit.py

fix(metadata): route strip output format by content, not extension

2026-07-14 17:16:36 +03:00

pill_gate_audit.py

Add full-surface verification harnesses and corpus sweep plan

2026-07-20 08:14:50 -07:00

qwen_scrub_prototype.py

feat(invisible): add Qwen-Image img2img pipeline (--pipeline qwen)

2026-06-19 20:44:36 -07:00

real_examples_e2e.py

Register the Qwen 千问AI生成 visible text mark

2026-07-21 16:41:16 -07:00

render_pill_silhouette.py

feat(visible): capture-less AI生成 pill (#54), inpaint fallback, MI-GAN backend (#56)

2026-07-06 20:38:23 +03:00

render_vendor_silhouettes.py

Register RunningHub, Baidu, and LibLibAI visible marks; park Qingyan and MiniMax (measured)

2026-07-22 13:03:03 -07:00

resource_ceilings.py

Report failed writes instead of crashing, and reject directories at parse time

2026-07-20 22:45:02 -07:00

robustness_suite.py

Report failed writes instead of crashing, and reject directories at parse time

2026-07-20 22:45:02 -07:00

scan_dataset.py

Add standalone dataset metadata/forensics scanner script

2026-07-22 17:53:02 -07:00

sidecar_regression.py

Add full-surface verification harnesses and corpus sweep plan

2026-07-20 08:14:50 -07:00

smoke_matrix.py

Add full-surface verification harnesses and corpus sweep plan

2026-07-20 08:14:50 -07:00

synthid_corpus.py

fix(scripts): drop rich import from analysis scripts (red CI after rich removal)

2026-05-31 15:41:50 -07:00

synthid_pixel_probe.py

fix(scripts): drop rich import from analysis scripts (red CI after rich removal)

2026-05-31 15:41:50 -07:00

vendor_cohort_harvest.py

Register the Qwen 千问AI生成 visible text mark

2026-07-21 16:41:16 -07:00

vendor_mark_calibrate.py

Register RunningHub, Baidu, and LibLibAI visible marks; park Qingyan and MiniMax (measured)

2026-07-22 13:03:03 -07:00

visible_alpha_solve.py

feat(visible): Samsung Galaxy AI mark removal (bottom-left reverse-alpha, #37)

2026-06-05 10:27:44 -07:00

visible_eval.py

Remove assume_ai, add tophat front-end and rival margin, fix two CLI defects

2026-07-20 08:14:50 -07:00

visible_groundtruth.py

Remove assume_ai, add tophat front-end and rival margin, fix two CLI defects

2026-07-20 08:14:50 -07:00

visible_positives.py

Add full-surface verification harnesses and corpus sweep plan

2026-07-20 08:14:50 -07:00

visible_recall_sample.py

Remove assume_ai, add tophat front-end and rival margin, fix two CLI defects

2026-07-20 08:14:50 -07:00

visible_removal_audit.py

docs: note content-based strip routing + SDXL watermarker fixes

2026-07-14 17:16:56 +03:00

visible_sheets.py

Remove assume_ai, add tophat front-end and rival margin, fix two CLI defects

2026-07-20 08:14:50 -07:00