SBIX
SBIX Visual Web Map · 2,000,000 images analysed
initializing graph...
VisualRank · cross-platform similarity graph of the visual web
SBIX — The Map of
the Visual Web
SBIX shows where an image appeared first and how it spread across the internet.
Using VisualRank — a PageRank-like cross-platform similarity score — SBIX builds a graph connecting visually identical images across independent platforms,
mapping detected co-occurrences, and which ecosystems show the strongest visual similarity signal.
How to read this map
NodeA platform ecosystem (Wikimedia, NFT market, Archive...)
LineThe same image detected independently on two platforms
ClusterA family of copied images traced over time
ColorPlatform origin — Common Crawl, Wikimedia, NFT, Archive
Images with the strongest cross-platform signal
Most cross-platform image
40 copies
5 platforms · first detected Common Crawl
Investigate →
Oldest resurfacing
1996 · 2022
Internet Archive + NFT marketplace · 26 year gap
Investigate →
Strongest cross-platform signal
4 platforms
NFT artwork · highest cross-platform score in dataset
Investigate →
Cross-ecosystem anomaly
Web + NFT
paypal.com asset detected in NFT marketplace context
Investigate →
—— Click any cluster to explore the story of an image ——
Scan an image
Platform Hubs
Image Analysis
Cross-platform detections
Timeline
Share this story
Copied.
First detected
Earliest observed platform
Total copies
Platforms reached
Index ID
Observed platforms
Temporal order does not imply a copy sequence.
Does your image travel like this?
Drop any image — SBIX detects visual matches across the indexed platforms in seconds.
→ Scan free
🔍 Visual Investigations
Anomalies · Archive jumps · Cross-ecosystem patterns · Suspicious propagation · Auto-detected from graph
Live data
Graph · VR-A 1M
DB size1,001,737
Total edges3,536,575
Avg edges/node3.54
Non-dangling43.9%
Connected components55,858
Cross-platform clusters7,404
Largest cluster size433,924 (43.4%)
Useful graph density8.05%
VR top1 score883.65
VR median0.56
VR spread ratio 735× — power-law structure forming. Connected components: 40,000 total graph families.
CPP Signal
CPP_core_pct2.06%
CPP_core families7,404
CPP_core nodes20,591
CPP_edges1,184,756
Platform pairs14 active
Google Moment target>10%
CPP_core_pct = only real signal. Never use cpp_pct (13.25% includes mega-cluster noise).
Noise / Signal
Mega-cluster188,498 nodes
Mega share95.5% of raw CPP
SourceCommonCrawl CDN
MOAT score64.04 "fort"
MEGA_THRESHOLD500 → 5000 (VR-B)
CC without pre-filtering = dataset artifact, not real propagation. Fix: §17 VR-B roadmap.
VR-B Roadmap
Next runVR-B · filtered CC
Target dataset~360k (CC 42%)
CPP_core est.6–9%
Google Moment est.500k A+C sources
generated_at2026-03-11T00:25Z
AlgorithmVisualRank Phase A v0.3
Bottleneck: crawl source quality (A+C), not speed. CC without pre-filter = pure noise.
Share this image story
Share to 𝕏 / Twitter
𝕏 Open in Twitter
Copied!
Platform ecosystems · similarity bridges between platforms · click a hub to explore detection clusters
Cross-platform co-occurrence · explore how image families appear across ecosystems
Cross-platform detections through time · watch image families emerge from 1996 to today · replay with slider
Cross-Platform Detection Timeline
All families visible
1/—
1996
2026
— today
2026
VR-B · visualrank_report_vrb_v031_2m.json · 2026-03-15
SBIX maps visually identical images across independent platforms — each line is a platform bridge where the same image appeared independently. Click a hub to explore · click a family to inspect its story.
2,000,000 images indexed · 8,757,417 similarity links · 6,412 image families · 91 platform bridges
Most cross-platform image
despixelsetdeshommes.com
First detected in 2019. SBIX found the same image appearing independently on 5 different platforms.
40 copies 5 platforms
Strongest cross-platform signal
FISHI FISH | AMBASSADOR OF OBJECTS
NFT minted on TzAll in 2021. Found independently on 4 platforms — strongest cross-platform signal in this dataset.
Strongest signal 4 platforms
Oldest observed cluster
Register of Maryland
Archived by Internet Archive in 1996. The same image was detected on NFT marketplaces 26 years later.
Since 1996 36 copies
NFT cross-platform
"tired froggie" · FROGGIE SERIES
Visually identical occurrences independently detected on Objkt, Teia, and Wikimedia Commons.
3 platforms cross-platform
LIVE · VR-B · v23
2,000,000
Images indexed
8,757,417
Visual copies detected
6,412
Cross-platform clusters
15
Platforms connected
91
Platform bridges
1996
Oldest indexed image
53.53% CPP creative
Cross-platform signal
Visual Investigations · 6 real detection cases · click to expand
Platform Ecosystems
All
Image Bridges
Visually identical image families detected independently on both platforms — confirmed cross-platform co-occurrence
Top similarity clusters
Image families with the strongest independent cross-platform spread — ranked by VisualRank cross-platform score.
1st
FISHI FISH
13 copies · 4 platforms
★★★★★
2nd
tired froggie
14 copies · 3 platforms
★★★★☆
3rd
sOUND
15 copies · 3 platforms
★★★★☆
pERSPECTIVE
teia · 16 copies · 2 platforms
★★★★☆
5th
HIEROglyphs
12 copies · 3 platforms
★★★☆☆
Recent similarity clusters
Most recently matched cross-platform image families in the index
teia_803645
Teia + CC + Wikimedia · 12 identical occurrences
NEW
objkt_KT1B34_50
Objkt + Wikimedia · 11 identical occurrences
wiki_2507998
Wiki + CC + Teia · 11 identical occurrences
Index updated · 568,958 fingerprints
SBIX maps visually identical images across independent platforms — each line marks visually identical occurrences detected across platforms.
Data: Common Crawl · Wikimedia · Internet Archive · Teia · Objkt · TzAll · fxhash · Met · Europeana · DPLA · NASA · Smithsonian · Wellcome · Rarible
Cross-Platform Detection Timeline
Select a cluster to see its history
Observed platforms
Image Spotlight
Does your image appear in this map?
Scan any image — SBIX detects visual matches across the indexed platforms in under 1 second.
→ Scan free
Detection history
⊙ Trace Image
Scan your image Detect visual matches across the indexed platforms in under 1 second.
→ Scan free
Cross-platform detections
Detection timeline
This view shows temporal cross-platform detections of visually identical images. It does not imply a directional propagation path.
1
Scan free · find copies
2
Check pro · detection report
3
Certify legal · eIDAS proof
What you see
Platform node
Size = images indexed
Color = platform type
Visual similarity edge
Width = shared image families
Cross-platform co-occurrence
Image family (cluster)
Group of visually similar images
detected across platforms
Strong cross-platform signal
High visual similarity score
observed across multiple platforms
Platform size = image count
Bridge width = shared families
Particles = cross-platform similarity
SBIX scores how widely an image
has spread across independent platforms.
More platforms = stronger signal.
↗ Click any cluster to explore
the story of an image.
Hover edges to see platform bridges.