Overview

Vearch is a scalable system for deep learning vector search, and particularly it can works as an open source visual search engine.

Architecture

• Components

  Master, Router and PartitionServer

• Master

  When you crate database or space you must use this service , default port is 8817 when you create dabase is only create a scope associate user permissions.

  When you create space ,the master will select relatively idel machine to create partition , when you delete space the master notice the related machines to delete local partition.

  Responsible for the management of distributed configurations.

• Router

  Supports restful api.create , delete search and update ， also when write document it routing function to related machine , to save it , you can define your routing args default is _id , and merge multiple searching results to one result.

• PartitionServer (PS)

  Hosts document partitions, raft-based replication.

  Gamma`is the core vector search engine. It provides the ability of storing, indexing and retrieving the vectors and scalars.

Quick start

• Quickly build a distributed vector search system with Restful api, please see docs/Deploy.md.

• Quickly build a complete visual search system, which can support billion-scale images. The image retrieval plugin about object detection and feature extraction should be extra required, For more information, please refer to docs/Quickstart.md.

API

VisualSearchAPI

• docs/APIVisualSearch.md

LowLevelAPI

• docs/APILowLevel.md

License

Licensed under the Apache License, Version 2.0. For detail see LICENSE and NOTICE.