An article prepared by Tecnalia (Partner of the LAW-GAME Consortium)

Emotion detection in the LAW-GAME training process

Emotion detection is the process of recognizing, interpreting, and responding to the emotions of others. Emotion detection is a rapidly evolving subfield of Artificial Intelligence (AI) and Machine Learning (ML) that has the potential to revolutionize the way we train people. Human emotions are detected from different sources, such as body language, facial expressions, psychophysiological signals and even tone of voice.

What are the benefits of emotion detection in the LAW-GAME training process?

By tracking the emotional state of the trainees, trainers can get a better understanding of how the participants respond to the material and their overall level of engagement. This data can then be used to adjust the pace of the program or develop more effective training materials to better fit the emotional state of the participants as needed to ensure the best learning environment for everyone involved. This can help to reduce stress and anxiety, allowing the trainees to stay focused and engaged in the material. For example, if the AI detects that a learner is feeling overwhelmed, the trainer can adjust the material to better suit their needs. This could involve slowing down the pace of the lesson or providing additional support. On the other hand, if the AI detects that a learner is feeling bored, the trainer can adjust the material to make it more engaging.

Emotion detection can also provide feedback to the trainer about their own behaviour.

For example, if the AI detects that a trainer is displaying negative body language, the trainer can adjust their behaviour to be more positive. This can help create a more positive learning environment and improve the overall training experience.

Apart from the two benefits mentioned before, emotion recognition can also be used to assess the effectiveness of the training and to help identify areas of improvement in the training process.

How are we detecting emotion in the LAW-GAME project?

In the context of LAW-GAME goals, there is little research done about the influence of emotion in forensic settings. For example, Sambrano et al. (2021)[1] conducted a set of experiments on an interrogation scenario, where the participants had to read a scenario where a suspect was arrested and rated: i) the suspect’s guilt, and ii) the extent to which they would use several tactics to interview the suspect. The results were consistent with the feelings‐as‐information and cognitive‐appraisal theories of emotion (e.g., angry or happy participants would be more inclined to judge the suspect as guilty (judgement), while sad participants would show a stronger tendency to select benevolent interrogation tactics). That is why in LAWGAME we consider that integrating emotion status of the trainee into the metrics used for training assessment can help to understand better the trainee’s evolution (related to judgment, decision making, and information‐processing).

In LAW-GAME project, a series of training games are being developed. Specifically, the partners involved (TECNALIA, AiDEAS and University of Malta) will focus in detecting trainee emotions in a Virtual Reality interrogation scenario. Related to emotion identification, eight different emotions will be considered (joy, trust, fear, surprise, sadness, anticipation, anger and disgust).

Currently, we have performed an experiment measuring participants’ brain, cardiovascular and electrodermal activities with different devices, along with video recordings to analyse the body pose. This information will help to infer the relationship among emotions, psychophysiological signals and human stance, and in this way, gaining a more holistic understanding of their emotional state.

A sample of experiment organization for data capture is provided below:

Participants in experiments for data capture

Conclusions and further work in LAW-GAME project

Overall, emotion detection can be a valuable tool in the training process. By tracking the emotional state of trainees, trainers can adjust the program to fit the needs of the participants and ensure the best learning environment. By responding quickly to any questions or comments that arise, trainers can also provide more comprehensive and helpful feedback. Ultimately, the use of emotion detection can help to create a more successful training program for everyone involved, providing an accurate training assessment based in emotion detection.

For more information visit the LAW-GAME website at: https://lawgame-project.eu/

Stay tuned on the project’s latest updates by following our social media channels at:

LinkedIn: https://www.linkedin.com/company/law-game/

Twitter: https://twitter.com/LawGame_EUH2020

LAW-GAME has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 101021714.

References

[1] Sambrano, Deshawn & Masip, Jaume & Blandon-Gitlin, Iris. (2021). How emotions affect judgement and decision making in an interrogation scenario. Legal and Criminological Psychology. 26. 62-82. 10.1111/lcrp.12181.