/davis-matlab

A Benchmark Dataset and Evaluation Methodology for Video Object Segmentation

Primary LanguageMatlabMIT LicenseMIT

Introduction

Package containing the Matlab implementation of the code behind: The 2017 DAVIS Video Object Segmentation Challenge.

You can find the Python implementation here.

Code Installation

  1. Adapt the value of db_root_dir.m to point to the root dir where DAVIS is uncompressed in your system (contains folders Annotations and JPEGImages)
  2. Run the script startup.m to add the necessary paths and perform some checks.
  3. [If necessary] Recompile using the script build.m in case the startup script detects some files missing.

Code Usage

  • The script demos/demo_eval_multiple.m contains a demo of how annotations and results are stored in the case of multiple objects.
  • The script demos/demo_sweep.m contains a demo of how the dataset images and annotations are read (all functions in db_util).
  • The script measures/eval_result.m runs the evaluation for the selected measures on a certain subset of the dataset.
  • The three measures used in the evaluation are found in the folder measures.
  • The folder experiments contains the scripts used to generate all plots and tables in the paper. global_table.m might be the best point ot start.

Evaluate your technique

  • The script demos/demo_eval_your_method.m contains a demo of how to evaluate your method (call measures/eval_result.m).
  • Add your results in the folder $root_DAVIS\Results\Segmentations\480p, as the provided precomputed results, in a folder my_method
  • Run measures/eval_result.m on your technique: eval_result('my_method',{'J','F','T'}). (You can select which measures to use - You can skip T for fast computation)
  • Show your results as in experiments\global_table.m

Helper functions

Two helper scripts are found in the folder helpers:

  • combine_multiple_pngs: To combine different masks from each object into a single PNG.
  • create_submission_zip: To check and crate the ZIP file to submit to Codalab for the challenge.

Citation

Please cite DAVIS 2017 and DAVIS 2016 in your publications if it helps your research:

@article{Pont-Tuset_arXiv_2017,
  author    = {Jordi Pont-Tuset and Federico Perazzi and Sergi Caelles and
               Pablo Arbel\'aez and Alexander Sorkine-Hornung and Luc {Van Gool}},
  title     = {The 2017 DAVIS Challenge on Video Object Segmentation},
  journal   = {arXiv:1704.00675},
  year      = {2017}
}

@inproceedings{Perazzi_CVPR_2016,
  author    = {Federico Perazzi and
               Jordi Pont-Tuset and
               Brian McWilliams and
               Luc {Van Gool} and
               Markus Gross and
               Alexander Sorkine-Hornung},
  title     = {A Benchmark Dataset and Evaluation Methodology for Video Object Segmentation},
  booktitle = {The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
  year      = {2016}
}

Contacts