ePrints.FRI - University of Ljubljana, Faculty of Computer and Information Science

Evaluating photo aesthetics using machine learning

Domen Pogačnik and Robert Ravnik and Narvika Bovcon and Franc Solina (2012) Evaluating photo aesthetics using machine learning. In: Data Mining and Data Warehouses (SiKDD 2012), Information Society - IS 2012, 8-12 October 2012, Ljubljana.

This is the latest version of this item.

PDF - Published Version
Download (498Kb)


    In this paper we propose a method for au- tomatic assessment of aesthetic appeal of pho- tographs. We identify significant parameters that distinguish high quality photography from low quality snapshots. On the basis of these parameters, we defined calculable features for automatic assessment of photography aesthet- ics using machine learning methods. The cal- culation of features depends heavily on the identification of the subject in photographs. With the subject identified, we defined and im- plemented various features to analyze various aspects of a photograph. The features were tested on two datasets. First dataset was ob- tained from Flickr and manually labeled for evaluation. Second dataset was based on pho- tographs from DPChallenge portal where sub- jects were identified with a face detection algo- rithm. Both experiments showed some promis- ing results. In this article we specify the fea- tures which contribute to a successful classifi- cation of photographs, analyze their influence and discuss the results. In conclusion, we o↵er some suggestions for further research.

    Item Type: Conference or Workshop Item (Paper)
    Related URLs:
    URLURL Type
    http://www.cobiss.si/scripts/cobiss?command=search&base=50070&select=(id=9441108)Alternative location
    Institution: University of Ljubljana
    Department: Faculty of Computer and Information Science
    Divisions: Faculty of Computer and Information Science > Computer Vision Laboratory
    Item ID: 2510
    Date Deposited: 09 Apr 2014 08:40
    Last Modified: 09 Apr 2014 08:40
    URI: http://eprints.fri.uni-lj.si/id/eprint/2510

    Available Versions of this Item.

    Actions (login required)

    View Item