Los nuevos algoritmos de IA predicen las huelgas de los grupos deportivos con un 80 % de precisión

Los nuevos algoritmos pueden predecir las acciones en el juego de los jugadores de voleibol con más del 80%[{“>accuracy. Now the Cornell Laboratory for Intelligent Systems and Controls, which developed the algorithms, is collaborating with the Big Red hockey team to expand the research project’s applications.

The algorithms are unique in that they take a holistic approach to action anticipation, combining visual data – for example, where an athlete is located on the court – with information that is more implicit, like an athlete’s specific role on the team.

“Computer vision can interpret visual information such as jersey color and a player’s position or body posture,” said Silvia Ferrari, who led the research. She is the John Brancaccio Professor of Mechanical and Aerospace Engineering. “We still use that real-time information, but integrate hidden variables such as team strategy and player roles, things we as humans are able to infer because we’re experts at that particular context.”

Ferrari and doctoral students Junyi Dong and Qingze Huo trained the algorithms to infer hidden variables by watching games – the same way humans gain their sports knowledge. The algorithms used machine learning to extract data from videos of volleyball games and then used that data to help make predictions when shown a new set of games.

Algorithms developed in Cornell’s Laboratory for Clever Programs and Controls can predict the in-game actions of volleyball gamers with greater than 80% accuracy, and now the lab is collaborating with the Huge Crimson hockey staff to develop the analysis venture’s functions.

The outcomes have been revealed within the journal ACM Transactions on Clever Programs and Know-how on September 22, and present the algorithms can infer gamers’ roles – for instance, distinguishing a defense-passer from a blocker – with a median accuracy of almost 85%, and might predict a number of actions over a sequence of as much as 44 frames with a median accuracy of greater than 80%. The actions included spiking, setting, blocking, working, digging, squatting, standing, falling, and leaping.

Ferrari envisions groups utilizing the algorithms to higher put together for competitors by coaching them with present sport footage of an opponent and utilizing their predictive skills to follow particular performs and sport situations.

Ferrari has filed for a patent and is presently working with the Huge Crimson males’s hockey staff to additional develop the software program. Utilizing sport footage offered by the staff, Ferrari and her graduate college students, led by Frank Kim, are designing algorithms that autonomously determine gamers, actions, and sport situations. One aim of the venture is to assist annotate sport movie, which is a tedious job when carried out manually by staff workers members.

“Our program locations a serious emphasis on video evaluation and information know-how,” mentioned Ben Russell, director of hockey operations for the Cornell males’s staff. “We’re continually in search of methods to evolve as a training workers in an effort to higher serve our gamers. I used to be very impressed with the analysis Professor Ferrari and her college students have carried out to date. I consider that this venture has the potential to dramatically affect the best way groups examine and put together for competitors.”

Past sports activities, the flexibility to anticipate human actions bears nice potential for the way forward for human-machine interplay, in response to Ferrari. She mentioned that improved software program might help autonomous autos make higher choices, carry robots and people nearer collectively in warehouses, and might even make video video games extra fulfilling by enhancing the pc’s synthetic intelligence.

“People will not be as unpredictable because the machine studying algorithms are making them out to be proper now,” mentioned Ferrari, who can be affiliate dean for cross-campus engineering analysis, “as a result of should you really have in mind the entire content material, the entire contextual clues, and also you observe a bunch of individuals, you are able to do rather a lot higher at predicting what they’re going to do.”

Reference: “A Holistic Method for Position Inference and Motion Anticipation in Human Groups” by Junyi Dong, Qingze Huo and Silvia Ferrari, 22 September 2022, ACM Transactions on Clever Programs and Know-how.
DOI: 10.1145/3531230

The analysis was supported by the Workplace of Naval Analysis Code 311 and Code 351, and commercialization efforts are being supported by the Cornell Workplace of Know-how Licensing.