The perception of robots before and after movement

The perception of robots before and after movement

Join us for a public presentation of MCDS seed funding projects

By Melbourne Centre for Data Science

Date and time

Monday, November 18 · 2 - 3pm AEDT

Location

Melbourne Connect

700 Swanston Street Carlton, VIC 3053 Australia

About this event

  • Event lasts 1 hour

Join Melbourne Centre for Data Science for public presentations of recent seed funding projects.

Presentation: "The perception of robots before and after movement"

Date: Monday 18th November

Time: 2-3pm

Venue: Mezzanine (M13), Melbourne Connect, 700 Swanston St

When you see a robot, do you think it’s more like an animal? Or more like a toaster? Robots are a special type of object, capable of independent movement, while also being non-living. As human beings, we represent visual objects in a region of our brain that separates living from non-living objects, or animate from inanimate. While this division my seem straightforward, this separation is not so clear-cut, with some objects appearing closer to, or further away, from others. Vision scientists typically study this area of the brain, which is responsible for object identification, using static images, neglecting the dynamic aspects that contribute to object representations. In this project, we include video stimuli to investigate how motion, in addition to shape, influences object perception of familiar and unfamiliar objects. Specifically, we employ robots as less familiar objects that sit on the animate-inanimate border and are shape matched to biological counterparts. Robots are capable of self-initiated movement, with both fast and slow dynamics, and varying levels of predictability. We ask participants (n=19) to perform a triplet odd-one-out task, as well as rank objects for agency, humanness and intelligence, before and after watching videos of the objects in motion. Prior to video watching, participants rank humanoids close to biological dogs, in terms of agency, humanness and intelligence. After video watching, rankings of agency increase for all fast moving but not slow-moving robots. In conclusion, viewing objects in motion affects our internal representations of them. Our results reveal how humans may internally represent embodied artificial agents alongside animals, and how this may or may not be influenced by movement.

Researchers:

Dr Astrid Zeman - Melbourne Centre for Data Science

A/Prof Marta Garrido - Melbourne School of Psychological Sciences

Prof Anthony Burkitt - Chair of Bio-Signals and Bio-Systems, NeuroEngineering Lab

A/Prof Hamish Meffin - Biomedical Engineering


Tickets

Organized by