1.Making fencing more entertaining

We used tracking, image analysis, and machine learning to make fencing more exciting.

2.Developing technology useful for training

By analyzing the sensing data, player might be able to improve how his train, or even to come up with new tactics.

3.Beyond referee judgement

The fencing rules and techniques that are difficult to judge even by referees, let alone spectators, were explained/described using a newly designed language, and a judgement on a play was made in real time.

Data will be obtained from existing data. The database and analysis software that will give meaning to the data will be developed on an original basis.

Data will be obtained based on existing technology such as motion capture cameras gathering fencer motion data. However, for this project we are developing on an original basis the machine learning database and software for judging — using collected physical data — the techniques employed by a fencer and assessing the fencer’s physical state (weighted pressure/electromyography/pupil movement/heart rate, etc.).

What fencing gains through this project:

1. For the spectators: Will be able to accurately grasp what happens in a match.

2. For the fencers: What fencers gain by experience and intuition will become perceptible and will help make their training more scientific.

3. Match recording: It will be possible for a fencer to clearly and accurately record his/her own techniques and to see what kind of match he/she had.

4. Judgement methods: Uncertain decisions formerly based only on the naked eye will become more accurate.

We aim to officially employ this technology at the 2020 Tokyo Olympics and this project will progress further.