Kozminski Techblog

Artificial Intelligence impacts almost every field we can think of. The same things happens in the case of the sports world. From analyzing player performance to predicting game outcomes, AI is transforming the way we watch, play, experience or even think about sports. I want to draw attention to some of the most important ways that AI is affecting and assisting the sports industry.

   I would like to start with the topic of how AI helps to improve players or athletes in the training. Nowadays, almost every professional sports club uses special apps, which helps them to track the players performance during their training or in a match. By doing so, they can check all the players metrics and then decide in which field they still need to improve or what to do to avoid a potential injury. It also plays an important part, when a player comes back after an injury. Doctors and physiotherapists can track the pace of the recovery and then use indicators to choose the right loads and form of therapy.

   Second of all, AI plays a big role in refereeing, in some major sports. For example tennis, where AI-powered systems are being used to assist referees in making calls on close line calls, such as foot faults, and also to track the ball’s trajectory. Or in football, where Video Assistant Referee (VAR) occurs in most of the major leagues and tournaments, helping with offsides or tough calls. We have to remember that referee’s job is stressful and human errors occurs, but by using Artificial Intelligence, we minimize them and the controversy associated with it.

   And last but not least I wanted to mention, is the way AI is improving broadcasting of sport events. Artificial Intelligence is an amazing tool to collect all the necessary data, which then can be used by broadcasters to highlight key stats, patterns or even real-time predictions of results. Last year Microsoft introduced a tool called ,,Goal Probability” for Spanish football league, which is used to calculate the percentage chance of a goal based on distance between the ball and goal, a player’s line of sight, while shooting and their distance to the defenders. I think it’s a fantastic way to interact with viewers and a fresh way to watch a football broadcast.

Sources:

https://medium.com/@mygreatlearning/how-is-ai-used-in-sports-cdfbdd97ad82

https://www.vizrt.com/community/blog/5-ways-artificial-intelligence-and-machine-learning-will-make-sports-broadcasting-smarter

https://www.sportsbusinessjournal.com/Daily/Issues/2022/01/24/Technology/microsoftdevelops-goal-probability-metric-for-la-liga-broadcasts

https://www.laliga.com/en-GB/news/laliga-takes-pioneering-step-by-adding-advanced-near-real-time-goal-probability-graphics-to-its-broadcasts-thanks-to-microsoft-technology

   Have you ever been in a situation, where a random track starts playing on the radio, but you have no clue what is the name of the song? I can’t even imagine how many times that has happened to me or how many of my most favorite songs, I wouldn’t have known if the Shazam application didn’t exist. That made me think on how is that possible, that with the one tap the app is able to compare millions of songs and find the one you were looking for. In 2003 the cofounder of Shazam – Avery Li-Chun Wang published the research paper, in which he explained how it works. To put it in a nutshell audio fingerprinting is the answer.

   According to the research by Wang each file is fingerprinted. This is a process in which reproducible hash tokens are extracted. Whole process was analyzed by Trey Cooper Professional Musician and Web Developer. The fingerprint is shown as a spectrogram based on frequency and time. ,,For each section of an audio file, the strongest peaks are chosen and the spectrogram is reduced to a scatter plot. Through a process called combinatorial hashing, points on the scatter plot are chosen to be anchors that are linked to other points on the plot that occur after the anchor point during a window of time and frequency known as a target zone. Each anchor-point pair is stored in a table containing the frequency of the anchor, the frequency of the point, and the time between the anchor and the point known as a hash. This data is then linked to a table that contains the time between the anchor and the beginning of the audio file. Files in the database also have unique IDs that are used to retrieve more information about the file such as the song’s title and the artist’s name. Once we have all of the possible matches for the Shazam user’s recording, we need to find the time offset between the beginning of the Shazam user’s recording and the beginning of one of these possible matches from the database. This offset in timing can be calculated by subtracting the time of the anchor-point pair’s occurrence in the Shazam user’s recording from the matching hash’s time of occurrence in the audio file from Shazam’s database. If a significant amount of matching hashes have the same time offset, that song is determined to be a match”. In conclusion the song you are scanning with your phone is being fingerprinted, based on that your audio file is being compared to all the files in the data base and if the hashes match your song name and the author is being shown to you.

Sources:

https://www.ee.columbia.edu/~dpwe/papers/Wang03-shazam.pdf

https://medium.com/@treycoopermusic/how-shazam-works-d97135fb4582

Most Instagram users probably have noticed, a recent trend, while scrolling through Instagram, to be more specific, stories shared by their friends or even some celebrities with images looking almost exactly like them. At least in my case, it was basically every second post. Out of curiosity, I had to check what is this all about.

   All answers led to the same place, which was Lensa AI: Photo & Video Editor application, produced by Prisma Labs and released in 2018 in order to edit photos and videos. The app gained so much popularity in late November 2022 due to introducing a new feature ,,Magic Avatars”. According to the creators ,,Powerful AI will generate a ton of mind-blowing avatars based on your identity and various styles”. I have decided to try it myself and the outcome was various. After downloading the app and selecting the option ,,Magic Avatars”, we need to upload 10-20 of our photos. For the best possible results, we should import pictures with variety of backgrounds and facial expressions and should avoid group shots, covered face and monotonous pictures. Then we select a gender and head to the checkout. The application offers us 3 options: 50 unique avatars (5 variations of 10 styles) for 23,99zł, 100 unique avatars (10 variations of 10 styles) for 34,99zł and 200 unique avatars (20 variations of 10 styles) for 44,99zł. After the purchase, we get the notification that work is in progress and it shows us the estimated time. In my case it was something about 15 minutes and I had my avatars ready in 10 different styles, which where Super Hero, Stylish, Mystical, Sci-fi, Astronaut, Anime, Adventure, Cyborg, Cosmic and Rock Star. As I mentioned before the outcome varied, some of them were looking exactly like me, some were close, but in a few facial features were not right, for example, some of them had different shapes of the head, and some did not even look like human beings. Some of them looked like creatures with creepy faces and three arms. For sure the best one in my opinion was the Cosmic style, which had the most realistic face with almost every detail visible and beautiful colorful space in the background. I think this application is a lot of fun for people who are looking for their potential profile picture in the social media or just like to have their pictures edited in some cool and never seen before way.

   Google Maps is a mapping and navigation application, which is one of the most used by people worldwide. Basically, the app shows us how to get from point A to point B, by any possible form of transport like car, bike, train or even by foot. We can choose if we would like to get there as soon as possible, for example using toll roads, avoiding traffic jams or if time does not matter to us as much and we would like to save some money using national roads.

   Recently Google introduced to us a new feature on their app, which is eco-friendly routes. This function helps us to choose the most environmental friendly route based on fuel consumption. All we have to do is to set our destination, then open options in the Google Maps app and look for ,,Prefer fuel-efficient routes” option. After enabling that, we only need to set which type of fuel our car uses, and we’re done. According to Google, this feature is available in almost 40 countries in Europe including Poland and by using it we lower our expenses on fuel, which is expensive at the moment and reduce CO2 emissions like in the United States of America and Canada, where that system was already successful reducing half million metric tons of CO2, which compares to removing 100 thousand cars from the roads. This is not the first time when Google encourages people to go more eco-friendly. They already have complex cycling routes and paths for pedestrians, a system helping to get to our destination using public transport or an option showing us where the closest EV (electrical vehicle) charging stations are located. All of this is a part of the Google campaign for this year which is to make people consider their choices and choose those that are more eco-friendly.

   I think this a superb idea in an important issue. We have to remember that environmental problems are relevant to all of us and every initiative that tries to fix them does matter on a grand scale. I am looking forward to seeing more companies following the same path.

Sources:

https://polska.googleblog.com/2022/09/nowa-funkcja-wyznaczania-tras.html

   The most anticipated sports tournament in the world FIFA World Cup starts on this Sunday (20.11.2022). This is the most important football (soccer) tournament played every 4 years to emerge a new World Cup champion. The opening ceremony and the football game between the host of the World Cup- Qatar and their opponent Ecuador on Sunday will begin a 4 weeks battle of the best teams in the world. This year’s tournament is not like any before, starting in November in the middle of the football season. Normally it was around June only because the temperature in Qatar is too high in the summer. It is still around 30°C every day during World Cup. It will be also a debut for FIFA’s brand new semi-automatic offside technology.

After the great success of VAR (Video Assistant Referee) being used on FIFA World Cup 2018 in Russia and then followed by major football leagues FIFA decided to still develop various technologies, which may be implemented into use. One of them is semi-automatic offside technology, a VAR tool responsible for making way faster and more accurate offside decisions. The system automatically determines the player position at the exact moment of the ball being played towards him, but how is that even possible? It is thanks to this year’s official World Cup adidas ball Al Rihla and 12 tracking cameras. The ball has a special sensor mounted inside, which sends data 500 times per second to establish the exact moment of pass. This is way more accurate than basing it on the optical tracking by the assistant referee. Cameras are mounted underneath the roof of the stadium and are responsible for tracking the ball and 29 data points on each player’s body 50 times per second, which determines the player’s position on the pitch and if he is on the offside or not. After the referee call 3D animation is being created and it is shown to the fans to dispel the doubts. According to FIFA expectations the whole process will take about 25 seconds per decision, so it lowers the current average wait time almost by 45 seconds per decision.

   A lot of fans, even some of the footballers are complaining about technology killing the true beauty of the game, but the football world, like everyone else wants to keep up with the newest technology and apply it into use in order to reduce people’s mistakes and save some time as in the case of the semi-automatic offside technology.

Resources:

https://www.fifa.com/technical/media-releases/semi-automated-offside-technology-to-be-used-at-fifa-world-cup-2022-tm