Glossary · Search
Index (face search)
In a face search engine, the index is the stored collection of face embeddings and the page addresses they came from, built by crawling public web pages and running face detection on their images. A search compares one query embedding against the whole index. What is in the index decides what can be found; what is not crawled does not exist to the engine.
Updated November 6, 2026 · 2 min read
How is an index built?
- A crawler fetches publicly accessible pages and the images on them.
- Face detection finds faces in each image; images without faces are discarded.
- Each detected face is aligned and turned into a face embedding.
- The embedding is stored with the image URL, the page URL and a crawl date, in a structure built for fast nearest-neighbour lookup.
- Periodic recrawls add new pages and drop pages that have disappeared, with a lag that varies by engine.
What is in it and what is not?
| Content | Indexed? |
|---|---|
| Public web pages: news, blogs, forums, galleries, company sites | Yes |
| Public profile pages that search engines can crawl | Sometimes, depending on the platform's robots rules and the engine's policy |
| Private accounts, stories, messages, logged-in views | No |
| Video content | Only thumbnails and stills published as images |
| Images removed after the last crawl | Until the next recrawl |
How big are indexes?
Engines rarely publish verifiable numbers. Clearview AI has stated figures in the tens of billions of images for its law-enforcement product; consumer engines describe their index in general terms. Size matters twice: a bigger index finds more, and a bigger index produces more look-alikes, the doppelganger effect.
Why does the index matter legally?
Because building it means processing the faces of people who never asked. The GDPR treats those embeddings as biometric data, the EU AI Act bans untargeted scraping of facial images to build a database, and the cases against Clearview AI and PimEyes turn on exactly this step. Opt-out mechanisms work by blocking a face from the index or from results; the opt-out routes directory lists them.
Questions
If my photo is not found, is it not online?
No. It may be on a page the engine never crawled, behind a login, or too small to detect. Absence from an index is not absence from the web.
How often is an index refreshed?
Engines do not publish schedules. Expect days to months of lag between a page changing and the index reflecting it.
Sources
- 1.Google Search Central, How Search works: crawling and indexing (checked November 6, 2026)
- 2.Regulation (EU) 2024/1689 (AI Act), Article 5 (checked July 12, 2024)
Entry reviewed November 6, 2026. Found an error? Tell us.
Related entries
- Term Crawler Crawler defined: the program that fetches public web pages and images for indexing, how robots rules and logins limit it, and what that means for coverage.
- Term Public web Public web defined: the pages anyone can load without a login, the only part of the internet a face search engine can index, and what falls outside it.
- Term Face embedding Face embedding defined: the numeric vector a network computes from a face, why same-person vectors sit close, how it is compared, and why it is biometric data.
- Concept Building a web-scale face index How face search engines crawl the public web, store billions of embeddings for nearest-neighbour lookup, keep the index fresh, and decide what counts as public.
- Directory Opt-out routes for every face search engine Opt-out routes for 20 face and image search engines side by side: form or email, verification asked for, stated timeline, cost and link. Checked Nov 6, 2026.
Cite this entry
FaceSearch Wiki editors. "Index (face search)". FaceSearch Wiki, updated November 6, 2026. https://facesearch.wiki/glossary/face-search-index/