Skip to content
Code for BMVC 2018 paper 'Incremental Tube Construction for Human Action Detection'
Branch: master
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.
results
src
LICENSE
README.md Update README.md Apr 2, 2019
main.m

README.md

OJLA

Official code for BMVC 2018 paper 'Incremental Tube Construction for Human Action Detection' If you use this code, please cite this Incremental Tube Construction for Human Action Detection:

@inproceedings{Behl2018IncrementalTC,
  title={Incremental Tube Construction for Human Action Detection},
  author={Harkirat S. Behl and Michael Sapienza and Gurkirt Singh and Suman Saha and Fabio Cuzzolin and Philip H. S. Torr},
  booktitle={BMVC},
  year={2018}
}

Pre Requisites

STEP-1 Installing required libraries

The code for the tube construction algorithm is built in Matlab. Most of the functions used usually come with most Matlab distributions. Users might need to install Bioinformatics Toolbox, if it is not already present in the Matlab distribution. This is needed for the graphconncomp function.

STEP-2 Downloading and preparing the dataset

UCF-101: Download the images and ssd detections for UCF-101 from link. After downloading the dataset, extract it within the repository as it is.

TESTING

Run the file 'main.py'

python main.py

To visualize the results, make the flag 'DISPLAY = 1' or 'DISPLAY = 2' in file 'main.py'

TO-DO

to-do list

Contact

If you have any queries at all, or need help with understanding/using the code or any part of it, please do not hesitate to contact me (Harkirat) at harkiratbehl1@gmail.com .

You can’t perform that action at this time.