A newer version of the code is available at DAVIS 2017

A Benchmark Dataset and Evaluation Methodology for Video Object Segmentation (DAVIS)

Package containing helper functions for loading and evaluating DAVIS.

A Matlab version of the same package is also available.

Introduction

DAVIS (Densely Annotated VIdeo Segmentation), consists of fifty high quality, Full HD video sequences, spanning multiple occurrences of common video object segmentation challenges such as occlusions, motion-blur and appearance changes. Each video is accompanied by densely annotated, pixel-accurate and per-frame ground truth segmentation.

Citation

Please cite DAVIS in your publications if it helps your research:

`@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}
}`

Terms of Use

DAVIS is released under the BSD License [see LICENSE for details]

Dependencies

C++

• Boost.Python

Python

• Cython==0.24
• PyYAML==3.11
• argparse==1.2.1
• easydict==1.6
• future==0.15.2
• h5py==2.6.0
• matplotlib==1.5.1
• numpy==1.11.0
• prettytable==0.7.2
• scikit-image==0.12.3
• scipy==0.17.0

Installation

C++

1. ./configure.sh && make -C build/release

Python:

1. pip install virtualenv virtualenvwrapper
2. source /usr/local/bin/virtualenvwrapper.sh
3. mkvirtualenv davis
4. pip install -r python/requirements.txt
5. export PYTHONPATH=$(pwd)/python/lib
6. See ROOT/python/lib/davis/config.py for a list of available options

Documentation

See source code for documentation.

The directory is structured as follows:

• ROOT/cpp: Implementation and python wrapper of the temporal stability measure.

• ROOT/python/tools: contains scripts for evaluating segmentation.

  • eval.py : evaluate a technique and store results in HDF5 file
  • eval_view.py: read and display evaluation from HDF5.

• ROOT/python/experiments: contains several demonstrative examples.

• ROOT/python/lib/davis : library package contains helper functions for parsing and evaluating DAVIS

• ROOT/data :

  • get_davis.sh: download input images and annotations.
  • get_davis_cvpr2016_results.sh: download the CVPR 2016 submission results.

Contacts

• Federico Perazzi
• Jordi Pont-Tuset