Skip to content

nufer12/visual-search-code

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Visual-Search-Code

This repository contains the code for the submitted mauscript:

Ufer, N., Simon, M., Lang, and Ommer, B. (2021): Large-scale interactive retrieval in art collections usingmulti-style feature aggregation

image

Repository description

  • General structure:
    • api: php code for the rest-api client
    • apache2: config file examples for apache2
    • mysql: template for the mysql database
    • frontend: angular code for the frontend
    • search_backend: python code for the initialization and retrieval workers including the rertieval algorithm

Installation

All code tested on Ubuntu 18.04

Software

Install general dependencies (use requirements.txt):

  • Python 3.6
  • Python packages:
    • pytorch 1.2
    • torchvision 0.4
    • faiss-gpu 1.6
    • scipy, scikit-image, scikit-learn, opencv-python
    • cython, easydict, h5py, hdfdict, pillow, pandas, requests
  • PHP 7.2
  • MySQL 5.7
  • phpMyAdmin 4.9 (optional)
  • Node.js 8.9
  • Angular CLI 6.1
  • Apache 2.4

Instructions

Follow the readme in api to install/configure the rest api

  • install php
  • configure api with mysql database password, api secret and root path
  • copy api to desired location

Follow the readme in apache2 to install/configure the apache server

  • install and configure apache
  • start apache server

Follow readme in mysql to install/configure the database

  • install mysql-server and phpmyadmin
  • import database template (includes interface user: test with password: test)
  • set mysql database password

Follow the readme in frontend to install/configure the frontend

  • install node.js and angular
  • build the single page application
  • copy to desired location

Follow the reamde in search_backend

  • install python and packages using requirements.txt
  • adjust start_init_worker_template.sh and start_search_worker_template.sh
  • start initialization workers
  • start search workers

Hardware

The frontend and REST API can run on a server without GPU support and the visual search backend can run on another machine with a single GPU (we used NVIDIA Quadro P5000).

References

This repository also includes external code. In particular, we want to mention:

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published