This code is based on the paper Towards Semantic Fast-Forward and Stabilized Egocentric Video on the First International Workshop on Egocentric Perception, Interaction and Computing at European Conference on Computer Vision (EPIC@ECCV 2016). The goal of the program is to stabilize a fast-forward version of a video, using the dropped frames to reconstruct distorted images during the processing.

For more information, acess the link: http://www.verlab.dcc.ufmg.br/fast-forward-video-based-on-semantic-extraction.

Contact

Authors

Michel Melo da Silva - PhD student - UFMG - miche.nosp@m.lms@.nosp@m.dcc.u.nosp@m.fmg..nosp@m.com
Washington Luis de Souza Ramos - MSc student - UFMG - washi.nosp@m.ngto.nosp@m.n.ram.nosp@m.os@o.nosp@m.utloo.nosp@m.k.co.nosp@m.m
Jo�o Pedro Klock Ferreira - Undergraduate Student - UFMG - jpklo.nosp@m.ck@d.nosp@m.cc.uf.nosp@m.mg.b.nosp@m.r
Mario Fernando Montenegro Campos - Advisor - UFMG - mario.nosp@m.@dcc.nosp@m..ufmg.nosp@m..br
Erickson Rangel do Nascimento - Advisor - UFMG - erick.nosp@m.son@.nosp@m.dcc.u.nosp@m.fmg..nosp@m.br

Institution

Federal University of Minas Gerais (UFMG) Computer Science Department Belo Horizonte - Minas Gerais -Brazil

Laboratory

VeRLab: Vison and Robotic Laboratory http://www.verlab.dcc.ufmg.br

Program

Dependencies

OpenCV 2.4 _(Tested with 2.4.9 and 2.4.13)_
Armadillo 6 _(Tested with 6.600.5 – Catabolic Amalgamator)_
Boost 1 _(Tested with 1.54.0)_
Doxygen 1 _(for documentation - Tested with 1.8.12)_

Compiling

The program can be compiled using either cmake or qmake tools.

Into the project directory, run the following commands:

        user@computer:<project_path>: mkdir build 
        user@computer:<project_path>: cd build
        user@computer:<project_path/build>: cmake ..
        user@computer:<project_path/build>: make
        user@computer:<project_path/build>: ./VideoStabilization < Settings_file > [-h] [Range_min] [Range_max]

Exemples of usage

        user@computer:<project_path/build>: < Program_name > < Settings_file > [-h] [Range_min = 0 ] [Range_max = num_frames ]

Example 1: Run VideoStabilization using Help option.

        user@computer:<project_path/build>: ./VideoStabilization -h

Example 2: Run VideoStabilization in the Experiment_1 processing the whole video.

        user@computer:<project_path/build>: ./VideoStabilization Experiment_1.xml

Example 3: Run VideoStabilization in the Experiment_1 processing from the 150 frame until the last one.

        user@computer:<project_path/build>: ./VideoStabilization Experiment_1.xml 150

Example 4: Run VideoStabilization in the Experiment_1 processing from the 150 frame until the frame 490.

        user@computer:<project_path/build>: ./VideoStabilization Experiment_1.xml 150 490

Documentation

To generate the documentation of the project, run the following command into the project folder:

        user@computer:<project_path>: doxygen doc/doc_conf

The documentation will be generated in the <project_path>/doc folder.

Citation

If you are using it to academic purpose, please cite:

M. M. Silva, W. L. S. Ramos, J. P. K. Ferreira, M. F. M. Campos, E. R. Nascimento, Towards semantic fast-forward and stabilized egocentric videos, in: European Conference on Computer Vision Workshops, Springer International Publishing, Amsterdam, NL, 2016, pp. 557�571. doi:10.1007/978-3-319-46604-0_40.

Bibtex entry

{Silva2016, Title = {Towards Semantic Fast-Forward and Stabilized Egocentric Videos}, Author = {Silva, Michel Melo and Ramos, Washington Luis Souza and Ferreira,Joao Pedro Klock and Campos, Mario Fernando Montenegro and Nascimento, Erickson Rangel}, Editor = {Hua, Gang and J{\'e}gou, Herv{\'e}}, Pages = {557–571}, Publisher = {Springer International Publishing}, Year = {2016}, Address = {Cham}, Booktitle = {Computer Vision – ECCV 2016 Workshops: Amsterdam, The Netherlands, October 8-10 and 15-16, 2016, Proceedings, Part I}, Doi = {10.1007/978-3-319-46604-0_40}, ISBN = {978-3-319-46604-0}, Url = {http://dx.doi.org/10.1007/978-3-319-46604-0_40} }