Social robots can also be equipped with a range of vocal signals that support effective communication even without speech – conclude researchers at the HUN-REN-ELTE Comparative Ethology Research Group and the University of Debrecen.
During the investigation, experts looked for an answer to whether the regularity of animal sounds – which influences the sound that makes another animal approach or move away – could be transferred to artificial “machine” sounds.
In a study published in the journal Scientific Reports, volunteers' approach and avoidance responses were measured using an interactive online questionnaire, in which participants controlled an on-screen character based on the artificial sounds they heard: they approached or moved away from the source of the sound – read the announcement from the ELTE College of Science on Thursday.
Using a mouse or keyboard, participants were able to move the shape in the picture closer to or further away from the displayed sound source. The researchers represented the sound source as a masked object so that it did not affect responses on its own.
““We created stimuli that were triggered based on the emotional sounds of people and animals, ranging from simple sounds to more complex sounds from a biological point of view.” Marta Gaksi, principal investigator of the HUN-REN-ELTE Comparative Ethics Research Group, explained in the announcement.
The results showed that artificially generated short tones generally elicited approach, while louder tones elicited avoidance, regardless of their complexity.
The pitch of the voice had a more complex effect: for high-pitched sounds, the response depended on how loud they were. High-pitched tones elicited avoidance, while quiet, high-pitched tones elicited approach. In the case of deeper sounds, biological complexity was important: deep sounds with a vocal formant (a vocal property resulting from the filtering/amplification effect of certain frequencies in the vocal tract, which contributes to the hollowness of the sound) elicited avoidance, while deep sounds without a formant elicited approach, they wrote.
According to the findings of the Hungarian researchers, social robots in the human environment can also be equipped with a set of acoustic signals that effectively support communication even without speech.Thus, in situations where language interaction is not required, the operation of assistive robots can become simpler and independent of culture and language. “It is a great honour to be associated with this study,” added Beata Kurczuk, one of the study’s first authors and a researcher in the HUN-REN-ELTE Comparative Ethics Research Group.
