Glossary

Semantic photo search

Also called: AI text search, natural language photo search

Semantic photo search finds photos from a plain-language description — "first dance", "speeches", "blue dress" — instead of filenames or tags.

It works by turning both the image and the search phrase into vectors in a shared space, so "kids on the dance floor" can retrieve the right frames without anyone tagging them.

It complements face search: face search answers "which photos have me", semantic search answers "which photos show the cake cutting".

Pixeva includes AI text search across event galleries.

Last reviewed 2026-09-02 · Pixeva

See it in Pixeva

Create a free event gallery and try face search, camera FTP, and a live wall on real photos.