VERSION 2 COMMIT TO BE USED WITH ORNAMENTO PROJECT ***

VGG Image Search Engine (VISE) is a standalone tool to search a large collection of images using an image region as query.

VISE builds on the C++ codebase (relja_retrieval, Sep. 2014) developed by Relja Arandjelovic during his DPhil / Postdoc at the Visual Geometry Group, Department of Engineering Science, University of Oxford. The VISE project is developed and maintained by Abhishek Dutta.

This work is supported by EPSRC programme grant Seebibyte: Visual Search for the Era of Big Data (EP/M013774/1).

From a common source code base, this software is built to address the needs of the following two groups of users:

• Desktop user: All the user interactions (indexing, administration, query, etc) occur through a web browser based user interface. The main properties of VISE being used in this mode are:

  • no technical skill needed to install and run the software
  • single click installer for Linux, Mac and Windows platforms
  • intuitive user interface for all interactions with VISE: indexing, administration, query, etc.

• Cluster user: This groups of users will be more advanced users with the ability to run shell scripts, install libraries and compile C++ source code. Furthermore, the VISE software will run in a server environment (with large number of nodes and memory) to serve a large number of user queries using REST API. This mode is only supported for the Unix platform. All operations (e.g. indexing, administration, etc) except search engine query will be performed using unix shell scripts. A web browser based user interface (or REST API) will be available for search engine query operations.

Abhishek Dutta
18 June 2018