2021 Tokyo Olympics “secret weapon” exposed: AI is not only responsible for timekeeping, but also forcing these referees to lose their jobs

The AI ​​that made everyone “unemployed” is here again!


Source: wired

[Guide] The AI ​​that made programmers unemployed has extended the “magic catch” to the referee, and now even sports events are not let go. At the Tokyo Olympics that opened recently, a certain “watch maker” used its beach volleyball AI that it trained for four years.

Recently, the 2020 Summer Olympics finally kicked off in 2021.


At present, my country’s Olympic athletes have won 6 gold medals, 2 silver medals, and 5 bronze medals, ranking first with a total of 13 medals.

In the beach volleyball game, the players also performed very well.

Take off, block, smash…


Every movement of a volleyball player on the court will be recorded.

Does the mediocre camera in all directions only have the video playback function?

You are wrong to think so!

The cameras on the stadium now use computer vision technology, which combines the cameras with artificial intelligence to provide real-time data of various games.

Artificial intelligence everywhere

In this Olympics, AI can be said to be everywhere.

Among the various AI applications, the most interesting may be the old Olympic partner Omega.

Omega spent 4 years training the AI ​​and used it in beach volleyball matches.

The volleyball project, whether it is model training or application, is not very simple.

There are many players and the ball is fast. It is difficult to judge the score of a ball with the naked eye.

In indoor volleyball, there are already “Eagle Eyes” assistance.

The Hawkeye system is also called the instant replay system. It is a powerful tool for the accuracy of the referee’s judgment . 

Of course, our director’s eyes are more powerful than eagle eyes…


And now, Omega makes beach volleyball also use AI.

For beach volleyball, Omega needs to train AI to recognize countless types of shots due to the differences in player positions and playing techniques.

This requires that the clothes worn by the athletes are equipped with gyroscope sensors, which can collect and analyze about 2000 sets of data per second.

These data information include the change of the spiking method, the type of pass, the flight path of the ball, etc.


After obtaining the raw data such as the player’s speed, direction of movement, and jumping height, AI can infer the corresponding type of shot and pass.

After combining images and motion, computer vision-based AI can effectively track and predict the trajectory of the ball.

It doesn’t matter even if the ball flies out of the camera screen, because AI can automatically fill in the missing part based on the data when the ball disappears and when it reappears.


The speed of a ball can be obtained in real time during the game 

From “artificial” intelligence to artificial intelligence

Before the rapid development of science and technology and the emergence of artificial intelligence, the time of the Olympic Games was measured using the most primitive method: naked eye + stopwatch.

At that time, Omega provided a watchmaker and 30 chronographs for the event. These pocket watches were certified by the Observatory at Neuchâtel in Switzerland as high-precision split-second chronograph pocket watches.


Omega high-precision split-second chronograph pocket watch certified by the Neuchâtel Observatory in Switzerland

Omega’s timing tools can accurately time to close to 1/10 of a second.

Beginning in 1948, Omega launched the first photocell terminal camera “Magic Eye” at the London Olympics.

The photocell can accurately record the moment the athlete crosses the finish line, which solves the accuracy problem caused by the elasticity of the finish belt for a long time.


It can be recorded for you when it is almost deformed

It was not until the 1968 Olympic Games in Mexico that it was completely converted to electronic timekeeping. At that time, the finish camera recorded 10 new world records.


The finish photo printer used in the 1968 Mexico Olympics, it was not clear whether it was a track and field athlete or a skateboarder.

In the 21st century, “faster, higher, stronger”, artificial intelligence has also contributed its own strength.

Artificial intelligence scores gymnastics

In addition to being applied to beach volleyball, AI will also be applied to various sports events.

For example, when swimming, AI is used for image recognition to calculate the number of strokes, real-time speed, and distance between players.

Or record the athlete’s running posture, analyze and improve the running posture, so that the athlete can run faster.


Sports biomechanics expert analyzes Su Bingtian’s running posture during training

In gymnastics, AI is also applied to detect athletes’ postures and examine athletes’ technical movements.


When the athlete flips and jumps, the jury will add or deduct points based on the exact position of the gymnast’s body.

Referees need to closely evaluate the angles of the gymnast’s joints, such as knees and elbows, but sometimes it is difficult for the human eye to see the slightest imperfections in the momentary movement changes, such as accurately measuring the bending angle of the joints.


(Upper left) camera screen, (lower left) timeline, (middle) player movements captured by AI, (right) “viewpoints of referees” for each technique

In the 2019 World Championships, the International Gymnastics Association applied the lidar technology developed by Fujitsu .

They study the potential of 3D sensing technology in sports and connect lidar with AI systems.

The 3D shape of the person is captured by irradiating the athlete with laser light about 2 million times or more per second and calculating the distance based on the reflection time.

In addition, an algorithm was developed to obtain 3D data of joint positions based on the acquired 3D data of athletes, and then score against the database.


Pommel Horse Technical Judgment Mechanism

The referee must not only judge whether the posture is accurate, but also look at the quality and performance of the performance based on the difficulty of the action.

Therefore, in the case of increased difficulty and faster action speed, the AI ​​automatic scoring system can provide referees with reference opinions and reduce the burden of referees.


Scoring support system

It seems that AI is not only limited to making programmers “unemployed”, is it also necessary to make human referees “unemployed” according to this posture?

Reference materials:


Posted by:CoinYuppie,Reprinted with attribution to:https://coinyuppie.com/2021-tokyo-olympics-secret-weapon-exposed-ai-is-not-only-responsible-for-timekeeping-but-also-forcing-these-referees-to-lose-their-jobs/
Coinyuppie is an open information publishing platform, all information provided is not related to the views and positions of coinyuppie, and does not constitute any investment and financial advice. Users are expected to carefully screen and prevent risks.

Like (0)
Donate Buy me a coffee Buy me a coffee
Previous 2021-07-26 04:58
Next 2021-07-26 05:01

Related articles