ALBAWABA - Dr. Hamdi Dibeklioğlu, an Associate Professor at the university's Department of Computer Engineering, highlighted the global trend of affective computing and discussed their efforts in using AI to examine various behavioral aspects, including facial expressions, speech tones, and postures.
Dibeklioğlu emphasized the growing popularity of AI models, particularly ChatGPT, in recent studies related to affective computing.
The team's latest endeavors involve the development of algorithms designed to detect depression levels through AI. Drawing parallels between AI-assisted analysis and traditional clinical observations, Dibeklioğlu explained how the system processes multiple data points to gauge the severity of depression during a session with an expert.
The study integrates clinical data for depression detection, with the researchers collaborating closely with psychiatrists. Dibeklioğlu underscored the importance of understanding the interplay between behavior and depression levels, citing patterns that align with established psychological theories. The AI system's role in supporting experts during sessions aims to enhance the overall diagnostic process.
In addition to depression detection, Dibeklioğlu shed light on another project focused on using AI to assess the degree of truthfulness in statements. The team examines various data, including sentence structures, tone of voice, and gazes, to determine the likelihood of deception. Dibeklioğlu emphasized the necessity of accurate algorithm training for successful lie detection with AI.
The researchers are also actively involved in personality detection using AI, evaluating different dimensions of personality through visual and auditory cues. Dibeklioğlu highlighted the increasing use of AI in personality analysis globally, with companies leveraging AI in job interviews to extract personality statistics.
Furthermore, the team is exploring the application of AI in pain level detection, particularly in adjusting medication doses. The potential impact of this AI-driven technology is significant, especially in scenarios involving children and infants where direct communication about pain levels may not be feasible.
Dibeklioğlu concluded by emphasizing the importance of ethical considerations in AI applications related to human behavior analysis. While AI offers valuable insights, he cautioned against relying entirely on AI for decision-making without human oversight. The team's goal is to leverage AI as a supportive tool while ensuring ethical and responsible integration into various aspects of life.
