Font- Font+ Italiano / English

Karate dataset

The Karate dataset contains full-body movements of athletes performing katas.
Data includes 3D positions of 25 body joints at 250 Hz. The data are in .tsv format.
The Qualisys optical Motion Capture System was used.

Seven athletes took part in the study. All of them received a martial art education in karate, however they have different levels of expertise:
  • 2 athletes are considered to be highly experienced (one of them is a karate trainer and the other one participated in several championships),
  • 2 other athletes have around 10 years of experience and
  • 3 athletes have around 5 years of practicing.

Each athlete was asked to perform the 2 katas two times: Bassai Dai and Heian Yondan.
In total 28 trials were obtained. Athletes perform katas with their own speed and rhythm.
The duration of the recordings for kata Bassai Dai is from 71s to 115s. For Heian Yondan it is from 50s to 97s.

Examples of the visualization of the MoCap data can be seen below (two synchronized points of view):

Karate dataset: example 1

Karate dataset: example 2

The whole dataset is available here:

Participant Bassai Dai Heian Yondan

The katas were ranked by 14 experts using the following questionnaire:

Q1) What is the overall quality of the performance?
Q2) How good is the performance in terms of stability of the body (e.g., lack of parasitic movement, wobbliness, disequilibrium)?
Q3) How good is the performance in terms of posture (e.g., alignment of the body)?
Q4) How good is the performance in terms of kick/punch power?
Q5) How good is the performance in terms of Kime?
Q6) How good is the performance in terms of rhythm within the kata (i.e., alternation/transition between quick and slow parts)?
Q7) How good is the performance in terms of intrapersonal coordination (i.e., synchronization of different parts of one’s body)?

For the ranking please check our paper.

If you have used our dataset in your research please cite the work:
Niewiadomski, R., Kolykhalova, K., Piana, S., Alborno, P., Volpe, G., Camurri, A., Analysis of Movement Quality in Full-Body Physical Activities, ACM Transaction on Interactive Intelligent Systems.
doi: 10.1145/3132369