{% extends 'retrieval/base.html' %} {% block bodyclass %}dashboard{% endblock %} {% comment %} ########################################################################### # # # Copyright (c) 2011 - Idiap Research Institute # # http://www.idiap.ch/ # # # # See file COPYING for the license associated to this software # # # ########################################################################### {% endcomment %} {% block content %} {% if user.is_authenticated %} {% endif %}
 

Copyright (c) 2012


  • HEAT is a content-based image retrieval application.
  • The application software is released under the AGPL v3.0 open-source license by N. Suditu, Idiap Research Institute. A brief documentation is available on-line.
  • N. Suditu and F. Fleuret, "HEAT: Iterative Relevance Feedback with One Million Images", in Proceedings of the IEEE International Conference on Computer Vision (ICCV), November 2011.
  • N. Suditu and F. Fleuret, "Iterative Relevance Feedback with Adaptive Exploration/Exploitation Trade-off", in Proceedings of the ACM International Conference on Information and Knowledge Management (CIKM), October 2012.

 
 

How the system works?


  • The system initializes by itself without any user input.
  • Searching is an iterative process. At each iteration, you have to click on the image that is the most similar to what you search for.
  • In a few iterations, the system aims to display images relevant to what you search for.

 


 

What the test is about?


  • You have to search for images relevant to the query written on top of the page.
  • As soon as you are satisfied with one of the images, stop searching and click on the 'One image is relevant!' button.
  • The test proceeds automatically to the next query in two cases: when you click on the 'One image is relevant!' button or when the searching exceeds {{ iter_max }} iterations.

 
{% endblock %}