Code and data related to analysis of object sketches at a semantic part-level
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
160-pp
annotated-final
results
LICENSE
README.md

README.md

sketch-object-part-analysis

The documentation below pertains to the approach presented in our ACMMM-2016 paper for analyzing part importance of category-epitomes, a novel visual representation for sketches developed by us.

Project Page : http://val.serc.iisc.ernet.in/eotd/epipartviz/

You can access our ACMMM-2016 paper 'Analyzing structural characteristics of object category representations from their semantic-part distributions' here.

Our ACMMM-2015 paper which introduces category-epitome, 'Eye of the Dragon : Exploring discriminatively minimalist sketch-based abstractions for object categories' can be accessed here.

Citation

Please cite our paper in your publications if you use our dataset or if it helps your research.

  @article{DBLP:journals/corr/Sarvadevabhatla15b,
  author    = {Ravi Kiran Sarvadevabhatla and R. Venkatesh Babu},
  title     = {Analyzing structural characteristics of object category representations from their semantic-part distributions},
  journal   = {CoRR},
  volume    = {abs/1509.04399},
  year      = {2015},
  url       = {http://arxiv.org/abs/1509.04399},
  timestamp = {Thu, 01 Oct 2015 14:28:48 +0200},
  biburl    = {http://dblp.uni-trier.de/rec/bib/journals/corr/Sarvadevabhatla15b},
  bibsource = {dblp computer science bibliography, http://dblp.org}
  }

Instructions for running (MATLAB) code

To obtain the relative importance of semantic-parts in a given category, assuming a particular sketch stroke order, run the following MATLAB script:

get_epitome_part_stats_driver

Modify the above script by following the comments in the corresponding .m file (get_epitome_part_stats_driver.m)

To obtain a numerical characterization of semantic-part distribution histogram for a given sketch stroke ordering, run part_distribution_stats.m

Example :

part_distribution_stats('results','temporal')

Additional information and Datasets

annotated-final : This is the directory containing the final set of sketch part contour annotations for the 13 object categories-original

results : This directory contains the per-category histogram of relative part importances for different sketch stroke orderings

160-pp/annotated-images : This directory contains the guideline images used by annotators while annotating sketch part contours

160-pp/categories-original : This directory contains selected object category images from the TU-Berlin sketch database (http://cybertron.cg.tu-berlin.de/eitz/projects/classifysketch/sketches_png.zip)

160-pp/categories-part-names : This directory contains part names for each category. It is for these parts that contour annotations are added on a per-object category basis.

Contact Ravi Kiran (ravika@gmail.com) in case you have questions, comments or suggestions.