- Main
Own Body Perception as Bayesian Causal Inference
- Samad, Majed
- Advisor(s): Shams, Ladan
Abstract
This dissertation investigates the principles of multisensory integration that underlie the perception of ownership over one's body. To that end, three experimental approaches have been utilized: 1) an investigation of the rubber hand illusion from the perspective of Bayesian causal inference operating in peripersonal space has indicated that this phenomenon is governed by the same principles of statistical inference that govern perception of external objects, 2) an investigation of the same model formulated to operate in the somatotopic space -- that which lines the surface of the body -- revealed that the integration of visual and tactile representations is again governed by the same process of causal inference, and 3) an investigation of the malleability of the somatotopic space has revealed that brief exposure to synchronous visual-tactile pairs at different locations can cause a recalibration of that space. In combination, these three investigations have made use of the Bayesian causal inference model that has been implemented in different ways, in order to model the respective spaces of relevance. Seeking to synthesize a complete account of body ownership, I then proceed to propose a unified account that makes use of the principles of Bayesian causal inference and performs a combined computation that operates in both somatotopic and peripersonal spaces in performing the inference as to which object is my body.
Main Content
Enter the password to open this PDF file:
-
-
-
-
-
-
-
-
-
-
-
-
-
-