Skip to content

Fully image-computable recurrent neural network for figure-ground segmentation

Notifications You must be signed in to change notification settings

Cavalletta98/FG_RNN

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

8 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

FG_RNN

Figure-Ground Organization with a Biologically Plausible Recurrent Neural Network

Demo

To run the model,

  1. Run Setup.m for instructions on how to run model.
  2. Model parameters can be adjusted in mfiles/makeDefaultParams.m
  3. A demo of the model is found in demo.m

Paper

For more details about the model and/or experiments, please see our paper:

@Article{Hu_etal19,
    Title = {Figure-Ground Organization in Natural Scenes: Performance of a Recurrent Neural Model Compared with Neurons of Area V2},
    Author = {Hu, Brian and von der Heydt, R{\"u}diger and Niebur, Ernst},
    Journal = {eNeuro},
    Year = {2019},
    Volume = {6},
    Number = {3},
    Doi = {10.1523/ENEURO.0479-18.2019},
    Publisher = {Society for Neuroscience},
}

Model Results and Evaluation

Model results can be found in the output directory. The results are separated by type:

  • edge (contour detection)
  • ori (figure-ground assignment)
  • group (segmentation)

Evaluation code for the contour detection task can be found here.

Evaluation code for the figure-ground assignment task can be found here.

We refer the reader to the authors' original papers detailing the datasets and benchmarks:

@Article{Arbeleaz_etal11,
  Author = {Arbelaez, Pablo and Maire, Michael and Fowlkes, Charless and Malik, Jitendra},
  Title = {Contour Detection and Hierarchical Image Segmentation},
  Journal = {IEEE Trans. Pattern Anal. Mach. Intell.},
  Volume = {33},
  Number = {5},
  Year = {2011},
  Pages = {898--916},
  doi = {10.1109/TPAMI.2010.161},
  Publisher = {IEEE Computer Society},
} 

@Inproceedings{Teo_etal15,
  Title={Fast 2D border ownership assignment},
  Author={Teo, Ching and Fermuller, Cornelia and Aloimonos, Yiannis},
  Booktitle={Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition},
  Pages={5117--5125},
  Year={2015}
}

Experimental Data

The data used to compare our model results with experimental results can be found here. If you use this data for your own research, please cite the following paper:

@Article{Williford_vonderHeydt16,
  Title={Figure-ground organization in visual cortex for natural scenes},
  Author={Williford, Jonathan R and von der Heydt, R{\"u}diger},
  Journal={eNeuro},
  Volume={3},
  Number={6},
  Pages={ENEURO--0127},
  Year={2016},
  Publisher={Society for Neuroscience}
}

About

Fully image-computable recurrent neural network for figure-ground segmentation

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • MATLAB 100.0%