AI decodes emotions in tennis players

Resume: Researchers have developed an AI model that can identify tennis players' emotional states with high accuracy by analyzing their body language during matches. The AI, trained on real-life images, can detect positive and negative emotions, although it is more adept at recognizing negative emotions. This technology has potential applications in sports training, healthcare and other areas, but raises ethical concerns about privacy and data misuse.

Key Facts:

• The AI ​​model accurately identifies the emotions of tennis players based on body language.
• Both AI and humans are better at recognizing negative emotions.
• Potential applications in sports, healthcare, etc., but ethical issues need to be addressed.

Source: KIT

For their research, Recognition affective states from the expressive behavior of tennis players Using convolutional neural networks, sports science, software development and computer scientists from KIT and the University of Duisburg-Essen developed a special AI model.

They used pattern recognition programs to analyze videos of tennis players recorded during actual matches.

Success rate of 68.9 percent

Our model can identify affective states with an accuracy of up to 68.9 percent, which is comparable and sometimes even superior to assessments by both human observers and previous automated methods, said Professor Darko Jekauc from KIT's Institute of Sports and Sports Science.

An important and unique feature of the research is that the project teams use real-life scenes rather than simulated or contrived situations to train their AI system. The researchers recorded video sequences of 15 tennis players in a specific setting, focusing on the body language displayed when a point was won or lost.

The videos showed players with cues including a lowered head, raised arms in elation, a drooping racket or differences in walking speed; these cues can be used to identify the player's affective states.

After being fed this data, the AI ​​learned to associate the body language signals with different affective responses and determine whether a point was won (positive body language) or lost (negative body language).

Training in natural contexts is an important step forward in identifying real emotional states, and enables predictions in real-world scenarios, Jekauc said.

People and machines recognize negative emotions better than positive ones

Not only does the research show that AI algorithms may in the future surpass human observers in their ability to identify emotions, it also revealed another interesting aspect: both humans and AI are better at recognizing negative emotions.

The reason could be that negative emotions are easier to identify because they are expressed in more obvious ways, Jekauc said.

Psychological theories suggest that humans are evolutionarily better adapted to perceive negative emotional expressions, for example because quickly defusing conflict situations is essential for social cohesion.

Ethical aspects need clarification before use

The research foresees a number of sports applications for reliable emotion recognition, such as improving training methods, team dynamics and performance, and preventing burnout. Other areas, including healthcare, education, customer service and car safety, could also benefit from reliable early detection of emotional states.

While this technology offers significant benefits, the potential risks associated with it must also be taken into account, especially those related to privacy and data misuse, Jekauc said.

Our research strictly adhered to existing ethical guidelines and data protection regulations. And in view of future applications of such technology in practice, it will be essential to clarify ethical and legal issues in advance.

About this AI and emotion research news

Author: Margaret Lehn
Source: KIT
Contact: Margarete Lehn – KIT
Image: The image is credited to Neuroscience News

Original research: Open access.
“Recognizing affective states from tennis players' expressive behavior using convolutional neural networks” by Justus Hartlieb et al. Knowledge-based systems

---

Abstract

Recognizing affective states from tennis players' expressive behavior using convolutional neural networks

This study describes an AI model using advanced Convolutional Neural Networks (CNNs) to recognize affective states in real sports environments, specifically tennis matches.

In contrast to previous studies that mainly used data obtained from actors and rudimentary statistical methods, the current study emphasizes the analysis of body expressions in real-world contexts, aiming for a more naturalistic representation of human emotions.

Our CNN-based models show an accuracy rate as high as 68.9%, outperforming or matching human observers in many cases. Intriguingly, both the machine learning models and human observers showed a shared tendency to identify negative affective states more effectively, which can be attributed to the more intense and straightforward expression of these states.

These results not only advance the state of the art in affective state recognition, but also pave the way for broader applications, including in healthcare and the automotive safety sector, thus representing a significant advance in the development of advanced and universally applicable emotional recognition systems.