Skip to content

Latest commit

 

History

History

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
 
 
 
 

README.md

THUMOS14 dataset

s This folder provides resources for evaluating action label predictions on videos from the THUMOS14 dataset. Most of the necessary code, including the evaluation script and ground truth labels, needs to be downloaded from the official THUMOS14 page.

This dataset is provided as supplementary material for the paper:

Open-vocabulary action localization with iterative visual prompting
Naoki Wake, Atsushi Kanehira, Kazuhiro Sasabuchi, Jun Takamatsu, Katsushi Ikeuchi (2025), IEEE Access, 5, 56908-56917

@article{wake2025open,
 author={Wake, Naoki and Kanehira, Atsushi and Sasabuchi, Kazuhiro and Takamatsu, Jun and Ikeuchi, Katsushi},
 journal={IEEE Access}, 
 title={Open-vocabulary action localization with iterative visual prompting},
 year={2025},
 volume={13},
 number={},
 pages={56908--56917},
 doi={10.1109/ACCESS.2025.3555167}}

The following is the citation for the THUMOS challenge, taken from the official THUMOS14 page:

@misc{THUMOS14,
  author = "Jiang, Y.-G. and Liu, J. and Roshan Zamir, A. and Toderici, G. and Laptev, I. and Shah, M. and Sukthankar, R.",
  title = "{THUMOS} Challenge: Action Recognition with a Large Number of Classes",
  howpublished = "\url{http://crcv.ucf.edu/THUMOS14/}",
  Year = {2014}}

Directory and File Structure

  • label_data_estimate_thumos14.txt
    This is an example file that contains estimated action labels. It is used as input to the evaluation script (see below). For details on the file format, please refer to the THUMOS14 challenge documentation.

Usage Instructions

  1. Download the THUMOS14 Evaluation Toolkit

    • As described in the THUMOS14 challenge documentation, download the evaluation toolkit (see "Section 4 Development kit").
    • Unzip the downloaded file (THUMOS14_evalkit_20140818.zip).
  2. Place the Label Data File

    • Move label_data_estimate_thumos14.txt into the THUMOS14_evalkit_20140818/TH14evalkit/results directory.
    • Install MATLAB or Octave, and from within the THUMOS14_evalkit_20140818/TH14evalkit directory, run the following command to compute the evaluation metrics:
    [pr_all, ap_all, map] = TH14evaldet('results/label_data_estimate_thumos14.txt', 'groundtruth', 'test')