Thursday, 10 February 2022

Dr. Darren Rhodes Research Seminar - Dynamic Bayesian Models of Time Perception. Friday 25th February, 2022. 1-2pm

Dr. Darren Rhodes Research Seminar - Dynamic Bayesian Models of Time Perception. Friday 25th February, 2022. 1-2pm

 

Psychology Department Research Seminar Series 2021-22

Dr. Darren Rhodes, Nottingham Trent University


Dynamic Bayesian Models of Time Perception

 


 


*** Everyone welcome.  No need to book in advance ***

Speaker: Dr. Darren Rhodes, Nottingham Trent University

Date:  Friday 25th February 2022

Time:  1-2pm

Location: 

https://mdx-ac-uk.zoom.us/j/94223657287?pwd=Q3I5Q2VzZmpNQklOd3hFdTN1Y0hMUT09 

 

 


Abstract:

The ability to decide when an event has happened and for how long it has lasted are cornerstones of perception and cognition. Surprisingly, no principled model exists that jointly accounts for perceptual performance in timing and duration. Here, we show how event-based and interval-based timing are interdependent by appealing to a unified computational mechanism. We propose that in sequences of events, estimates of timing and duration are made probabilistically following a Bayesian-optimal principle that combines the history of durations with the estimated of time points to predict when stimuli might occur. We conducted an experiment where we varied the reliability of stimuli in a sequence showing that perceptual judgments follow the mathematic predictions derived from a specific set of assumptions. These results support a unified, parsimonious and physiologically-plausible account of the perception of event timing and duration.

 

About the speaker: 

Dr. Darren Rhodes is a senior lecturer in computational neuroscience and open research at Nottingham Trent University. He previously completed his PhD at The University of Birmingham with Dr. Max Di Luca, before a post-doc at the Sackler Centre for Consciousness Science with Anil Seth, and as a visiting lecturer at the Norwegian Technical University in Trondheim, Norway.





No comments:

Post a Comment