Artificial Intelligence In Sports,AI Projects In Sports

Artificial Intelligence in Sports new AI which is attractive, safer and more profitable.

During the pandemic in the world, many people did not know where to go for fitness, the only way to exercise for many was to do the exercises at home.Neural networks quickly learned to recognize human movements and control the correctness of exercises.

Human motion recognition Ai

In this regard, several technology companies start developing a solution to replace gyms at home. They are working on the production of fitness mirrors to turn the house into a gym.

Prior to the pandemic, a particularly important and significant trend for young people was a passion for fitness, attending sports clubs and exercising. Well, this passionate fitness enthusiast is going crazy and forced to look for new alternatives in order to maintain a perfect figure.

A bit of history

The concept of the home gym is relatively recent, but it took off in 1982 when actress Jane Fonda began publishing aerobics workout videos that anyone could buy as a cassette.

Today, this trend has changed significantly – now training tips can be found on websites, in applications, or even on social networks.However, many prefer to work out in gyms, in group classes, but now the pandemic has made a change – you need to exercise at home, and it is very likely that this habit will continue in the future.

AI can receive data from various sources, such as player activity statistics, match video, or physical activity metrics, and analyze them. The results of such analysis can be used by coaches, players and managers to make effective decisions, and in some areas – to completely automate the decision-making process. The areas of application of AI in sports are becoming more and more. Let’s highlight the main areas of Artificial Intelligence in Sports.

Before the game: Preparation of athletes: – Nutrition – Physical preparation – Biomechanics (skill / technique) – Psychological preparation • Injury prevention • Strategy and tactics selection for the game • Team building
In-game: Game Analytics • Injury Recovery • Injury Prevention • Virtual Assistant Referees • Coach Analytics Services
After the game: Fan engagement • Content creation • Media rights • eSports • Betting
Scouting: Search for potential candidates to move to the team • Evaluation of the potential of athletes

Recording of a hockey game with Iceberg

Iceberg , a Canadian startup with Russian roots , is using Computer Vision algorithms to help coaches improve the performance of their teams. The service creates a detailed analysis of matches, where game events are collected and classified. This helps to build an optimal strategy for future matches, to notice the weaknesses and strengths of the players.

Iceberg automatically breaks down the game record for each metric collected, saving coaches time and allowing them to focus more on analyzing team interactions, understanding overall trends, and individual player performance.

The game can be recorded live – then three cameras are installed on the site, each of which films its own zone of the site. Game data from TV broadcasts can also be analyzed. After recording, a large array of players’ coordinates is collected and geopositioning analysis is done – on this basis, you can analyze speed indicators, movement around the court, for example, who worked out in defense and who did not.

Recognition and analytics of chess games with idChess

At tournaments and in the learning process, chess players keep a record of their games. idChess , a platform for recognition and analysis of chess games, does it for chess players. idChess recognizes the moves during the game and writes them down in chess notation for the player.

After the game is over, a game in gif or pgn format can be sent, for example, to a coach, or shared on social networks.

Recognition requires a phone with an installed application and a tripod holder to securely mount the phone above the chessboard.

Also, with the help of idChess, you can conduct online broadcasts of tournament games in real time. Suppose the children participating in the tournament are in the same room, and in the lobby their parents watch the progress of the tournament and the game of their children.

At the same time, idChess does not need access to the Internet – all recognition functions are available offline.

The application helps the player track his progress in chess and improve the quality of learning, analyze games. And for children who can’t write, idChess has already become an indispensable tool.

Motivation for athletes from HearMeCheer

There is another platform Hearmecheer introduced as a Artificial Intelligence in Sports.The coronavirus pandemic has put everyone in isolation. And although sports were allowed to resume, the presence of spectators in the stadiums was limited. To prevent athletes from playing matches in silence, Canadian startup HearMeCheer came up with a solution that allows players to hear the voices of fans.

HearMeCheer collects “crowd noise” from fans while watching a broadcast and distributes it to the stadium during games.

The applause of the fans, who, at the same time, are not at the stadium, but are sitting in front of TVs at home, is heard by the players on the field!

HearMeCheer automatically collects microphone data from fans at home and then sends this data to its servers, where the sounds are compiled into a single audio stream that can be played live during the broadcast or in the stadium.

In the future, more people will be able to attend the game than the stadium can accommodate.

Fast and Furious 15: Formula 1 and artificial intelligence

Formula 1 is a leader in the use of data analysis. During the races, not only the condition of the race cars is monitored, but every movement on the track is monitored: a large team monitors the race on the site itself, and hundreds more at the headquarters. Since it is impossible to analyze everything live without resorting to computer analysis, complex simulations of possible scenarios are carried out before and during each race . Strategists choose the perfect moment for a pit stop, looking for opportunities to overtake an opponent.

NASCAR races in the United States, before the introduction of AI, mechanically tracked violations, collecting information from video replays in parts using multiple monitors. This required dozens of trained employees, and the violations recorded by them could be challenged. Then they introduced AI – a Microsoft development based on Windows 10 , installed 46 Sony Ipela cameras on the track. It was possible to combine six categories of data on one screen – historically collected data, timing, scoring, referee’s decision, video from the track and car positions. The first test recorded about 100 violations.

Personal training with The Mirror

Regardless of the sport, AI services have a wealth of training information that allows coaches to analyze progress and set clearer goals for each player. The Mirror is a mirror that, with the help of a camera, recognizes movements during training and analyzes the condition of the athlete. The startup is selling a new type of device, a human-height display with webcams and speakers, priced at $1,395. During training, the user sees not only himself, but also the trainer’s record, and the automation evaluates the compliance of the movements with the pattern. Program algorithms complicate the level of training and provide feedback based on the user’s goals and preferences for maximum results. Mirror was launched in September 2018 and attracted almost $75 million in investment.

An easy way to get into big sport with the NBA

We see Artificial Intelligence in Sports of NBA.The NBA is using a new player recruiting app. It gives tasks through the application and analyzes their abilities with the help of AI. NBA Global Scout allows you to “pass the interview” and showcase your playing skills from anywhere in the world by recording multiple videos. It’s a great opportunity for the company not to miss out on talent in countries where basketball is not as popular. In March 2020, the app became one of the most downloaded in the sports category. At the beginning of March 2020, the application was tracking approximately 100,000 throws per day, and at the end of April 2020, it was already 600,000 throws.

AI paves the way for sports victory for everyone, athletes and fans alike, with real-time game statistics. For players and coaches with predictive game tactics, it allows you to choose the right strategy and even warn the player in the event of a potential decrease in performance or injury.

Mirrors with artificial intelligence

For this reason, several manufacturers of equipment have begun to develop artificial intelligence mirrors for training at home, which will allow users to attend virtual sports classes. The first such mirror, eponymously named “Mirror”, was created in 2018 – it allows fitness enthusiasts to explore yoga, cardio and Pilates.

The artificial intelligence or virtual simulator that we see in the mirror gradually increases the load on the user, adapts to the physical condition, tracks progress and recommends certain workouts that will bring results for user settings, for example, someone who wants to lose weight, someone to gain muscle mass.

In the near future, it is planned to be able to use it with various sports accessories, such as training with rubber expanders, hand trainers and will monitor heart rate and other body parameters.

The mirror in which the camera is mounted monitors the position of the body, and neural networks control the correctness of the exercise.

Such technologies can be used to teach people dance exercises, and control the correctness of the dance.

Artificial intelligence controls dancing

The AI can act as a robot coach while playing a sport or learning a new dance, but on a day-to-day, off-mode, it will serve as decoration.

How to apply artificial intelligence in sports ?

In big sports, there are also more and more areas of reconciliation every year, AI can determine body parts on video and track the execution of a blow, count and clearly perform the exercise.

More from Ai Industry :

5 amazing examples of how AI is making sports more attractive

1. Satisfi Labs chatbot helps guest navigate

How it works: Guests of any event usually have the same questions: from which side is it better to drive up to the building, where exactly is the paid place in the hall, what kind of food and at what price can be purchased before the start or during the break. Satisfi Labs has figured out how to simultaneously give hints and shoot the main questions for future visitors to sports events.

For example, if a viewer is interested in parking, artificial intelligence formulates key questions for him on this topic and determines the best place to leave the car. All questions require simple answers from the guest such as “ yes-no” or “ north-south”, and as a result, the bot names the parking zone from which it is closest to go to the place in the arena.

In parallel, the bot generates request statistics. For example, arena food outlets receive data on the prices of which dishes are most often interested in. In this regard, they can expand the range and optimize advertisements inside the stadium. And the arena management can learn how to improve navigation and what services, not yet implemented, are most often requested by visitors, and thus modernize on this basis.

How it will evolve: It took only a year and a half for Satisfi Labs to penetrate the largest US arenas ( Mercedes-Benz Stadium in Atlanta, Barclays Center in Brooklyn, MetLife Stadium in New Jersey), NHL clubs have already come for the technology ( Tampa Bay Lightnings) and NBA ( Oklahoma City). The company is actively expanding beyond sports – a key technical solution is versatile enough to work in other areas.

Assistants from Satisfi are also good because they can be easily integrated into existing partner platforms: smartphone applications, Facebook messenger or a separate page of the official website.

This opportunity is already being used, for example, by Hilton hotels. For them, this development is an opportunity to expand the boundaries of hospitality and create comfort even outside their premises: the service is used by guests who have arrived in an unfamiliar city. It is important for them, as well as for visitors to the sports arena, to know where it is better to have lunch or buy the right thing. For Hilton, such a bonus to the usual services is an additional chance to demonstrate care for the guest and remain in memory as a brand that creates comfort not only for an overnight stay, but also for staying in the city as a whole.

Scaling from the coverage of a sports arena to the whole city continues successfully. Intermediate point – the largest shopping centers. Like the Mall of America with 500 stores. In parallel, Satisfi teamed up with hologram specialists from VNTANA to create an assistant with a visual identity. Communication with him will have an even stronger wow effect, and ideally, the hologram will be able to recognize the faces of people who seek help. When contacted again, artificial intelligence will be able to use the experience of the previous conversation.

2. Social media highlights available in 5 minutes: IBM Watson project at Wimbledon and US Open tennis tournaments

How it works: The oldest Grand Slam tennis tournament , Wimbledon, together with partner IBM, made a complex technological breakthrough. In addition to working with big data, in 2017, Wimbledon began using the IBM Watson supercomputer to create cuts of the best moments of matches – highlights. This is a whole system based on the use of artificial intelligence.

Cognitive Highlights is a collaboration between IBM Research and IBM iX. It was first used in 2017 at the Masters Golf Tournament, it made it possible to watch the most interesting moments from different parts of the course.

Thanks to machine learning, the developers taught the computer to recognize the emotions of tennis players after the draw: a clenched fist, a sharply raised hand, a smile, a jump. To this was added an analysis of the reaction of the audience: disappointed and joyful sighs, applause. Also, the computer understands the significance of the moment and highlights the draws on break points, set points and match points, and also takes into account statistics – for example, the speed of delivery.

Based on this, Watson selects the key moments of the match and mounts them in a video for Wimbledon ‘s social networks. The system needs 30 minutes to collect an overview of the game, a person needs 45. This is especially important in the first days of the tournament, when all 19 courts are involved, many matches are taking place in parallel and it is problematic to keep track of all. In 2017, Cognitive Highlights was installed on six courts.

How it will evolve: Another Grand Slam tournament, the US Open, has already started using the same technology. And there, the time required to collect highlights was reduced to five minutes after the match.

Obviously, the developers are already close to the maximum speed. Their next steps could be proposals for the players themselves: in team sports, there have long been programs ( without artificial intelligence, marks are set manually) that allow, for example, a hockey coach after a match to quickly access only fragments of the game in the majority. Slices of all tennis right/left shots or serves can also be useful for game analysis.

3. Actual ” stories” appear right during the match: HEED opportunities for different sports

How it works: Founded in 2016, HEED is positioning itself as a product that changes the way you watch sports. HEED believes that the modern fan does not always have the opportunity and even the desire to watch the game in its entirety, but it is still important for him to stay in the subject.

Using IoT data analytics and artificial intelligence, the platform has developed a new way to provide fans with moments of sporting events in the form of short “ stories” right in the course of the match. It’s the same format that has spread across many apps following the success of short vertical videos on Snapchat and Instagram.

HEED currently has three main partners: Euroleague Basketball, the UFC’s major mixed martial arts league, and PBR ( Professional Bullriding, Rodeo). Since the 2017/18 season, each arena of a Euroleague participant has been equipped with special sensors developed using its own technologies that analyze what is happening in the arena: the behavior of spectators, the actions of players and coaches. This information is combined with the accumulated data and delivered to the application in the form of a video. It’s not just about highlights in real time, but about finding unique moments and data that can hook a fan.

For example, during the Khimki -Fenerbahce match, 10 short videos appeared on the HEED app, while none during the first quarter. These are different clips, ranging from dunks indicating the height of the jump of a basketball player and the time spent in the air, to measuring the noise of the fans in the decisive segment of the match.

How it will develop: The UFC example can be considered even more advanced: in addition to the already mentioned technologies for tracking movements, emotions and reactions of the viewer, HEED has achieved the introduction of sensors ( so far test) in athletes’ gloves. This way they get a lot of data: from the strength of the blow to indicators of the physical and emotional state.

Based on this information, it is possible to predict whether an athlete is able to recover from a blow or how long he will last in a fight. However, it is important for the UFC to maintain a balance between technology and fair refereeing, so until this data gets to the fans, only specialists analyze it.

The PBR bull riding series took a different path. HEED technologies open up new possibilities for analyzing rodeo, a very fast sport. Tracking the movements of the bull and the rider is not easy, and HEED data allows you to identify the position in space and each action. Artificial intelligence should actually become a judge who will evaluate each race objectively.

4. Copyright holders kill piracy :

How it works: Sports piracy takes a lot of money from event organizers, but it’s also a huge business in itself. The Mayweather-McGregor fight was the most pirated event in sports history. Illegal broadcasts could be found on all social networks and were tracked and blocked by Irdeto. Its representatives said that one of the 239 broadcasts was watched by 472 thousand people at a particular moment. Of course, this is a financial loss for the organizers of the fight.

Irdeto is constantly working to improve its piracy tracking systems. One of the newest developments is related to the use of artificial intelligence . A computer, in addition to existing technologies, can analyze the constituent parts of pictures: logos of sponsors of an event, faces, text on a graph.

How it will evolve: The Pirates are learning to deal with their barriers by obscuring or blurring logos and inserting billboards stylized as other TV channels. However, neural networks are always in the lead in this race: they make it possible to detect translation attributes that are not obvious at first glance in a complex.

Given the serious redistribution of influence in the broadcasting market and the rapid growth in the number of rights holders in the digital environment, which only makes it easier to illegally stream broadcasts of major sporting events, the fight against piracy will develop. Irdeto is already thinking about how to set more complex or even unsolvable tasks for pirates.

Irdeto continues to train artificial intelligence. The next step, according to the vice-president of the company, Peter Oggel, will be to discover the colors of the uniforms of the playing teams – for example, Barcelona. Pirates will definitely not repaint T-shirts.

5. Auto racing will become cheaper and safer: Microsoft’s solution for NASCAR

How it works: The NASCAR Series, popular in the United States touring oval track races, has long had a problem: riders who violate the rules when refueling or changing tires get a significant advantage. At first glance, crossing the boundary lines is a minor offense. However, in a race where the winner’s lead can be less than 0.001 seconds, every detail is very important. And the intersection makes it possible, for example, to slow down later or accelerate earlier.

Before the introduction of artificial intelligence, such violations were tracked by the organizers’ employees: there were about 50 of them at each race, each monitored their section through the monitor, recorded violations, and then double-checked them on repeat. Now the organizers need 16 people who process the data of the new system developed with the participation of Microsoft based on Windows 10.

Instead of 50 people, the organizers installed 46 Sony Ipela cameras with 8-80 mm lenses from Fuji. With the help of trigonometric formulas, an accurate model of the route is obtained, data from the cameras are superimposed on it, and as a result, violations can be undeniably determined. Teams and riders have no chance to protest: everything is recorded and documented.

During test races, when the development was already used in beta mode, but the teams did not notify about it, it recorded up to 100 violations. Already after the official introduction in May 2017, the first race was held, in which there was not a single violation.

Neural networks are being used for another innovation: Ford, which has invested about $1 billion in Argo AI, is using its developments to monitor the condition of the car. Previously, tire wear and external damage were tried to be controlled by eye. However, at high speed and in a situation where many cars often change color due to short-term sponsorship agreements, this is not very effective.

The neural network does a better job: it captures the position of the desired car on the track during a NASCAR race, takes a picture of it, and instantly matches it with an image of a car in perfect condition. This makes racing much safer: the outside team can accurately monitor wear and tear and call the pilot to pit before he encounters the difficulties of piloting.

How it will evolve: The technology, which was invested in at one time, has already saved several million dollars for NASCAR. The organizers did not disclose at what point they planned or plan to reach the break-even point, but this will inevitably happen.

First of all, the costs have fallen, since fewer employees are required at each stage, and the installation of equipment and workplaces is easier. Now some of them are inside a large truck filled with equipment – it moves from one track to another and just requires free space and connection. In addition to development, Microsoft only needs 10 people who stand right in the pit lane next to the teams and control other types of violations.

In a NASCAR season of 38 races, the amount of savings is growing almost every week. Many racing series, including Formula 1, are taking care of cost reduction. However, the way of the majority is to limit the cost of teams to invest in the development or testing of new products. NASCAR did the wiser thing: it brought in artificial intelligence to overhaul a costly internal process.

The next step is to train the system and transfer other data collected using the same cameras to the teams and race participants. It is planned that deeper analytics will help them in development. In addition, data from Microsoft can make race broadcasts more attractive, work in this direction is also already underway.

Ford’s neural network training project involves extensive collaboration, with the automaker’s investment primarily needed to develop self-driving cars. Helping NASCAR is just a small part of the project.

Conclusions

In conclusion, the following should be said. Artificial intelligence itself is still in its infancy. However, even today it is possible and necessary to use innovative developments, including applying developments in fitness or big-time sports.

Today there are no separate sports for robots based on artificial intelligence, but it is likely that there will soon be separate competitions in which only robots participate, with built-in neural networks that will compete with each other.