Architecture

Vertx technology is based on extracting digital signatures from audio and video content (called digital fingerprints) and adding them to a searchable index. Much likely like human fingerprints, digital fingerprints uniquely identify a chunk of audio/video content. Fingerprints are highly compressed when compared to the original media content, and it is not possible to restore the original content from digital fingerprints.

The three main components of the Vertx ecosystem are Vertx Cloud, Vertx Backend and Vertx Clients.

Vertx Architecture

Vertx Cloud stores indices of the reference content and serves search and registration requests from Vertx Backend.

Vertx Clients can upload media content for fingerprint extraction to the Vertx Backend servers through the REST API.

When using Vertx command line utility or container interface, fingerprint extraction happens on the client side and no original media content is being uploaded. Only extracted fingerprints are sent to the Vertx Backend servers for search or registration.

Metadata for all matched reference content is returned to the client.

All the reference content in the Vertx cloud is organized into buckets. Buckets can be private or public. By default, all search requests are sent to public bucket. It contains fingerprints for millions of songs and thousands of movies. Private buckets contain content registered by a particular customer (owner) and can only be accessed and managed by the owner.