Ward, Ryan 
ORCID: 0000-0002-9850-5191
(2023)
Virtual Synaesthesia: Crossmodal Correspondences and Synesthetic Experiences.
PhD thesis, University of Liverpool.
|
Text
201366711_Mar2023.pdf - Author Accepted Manuscript Download (10MB) | Preview |
Abstract
As technology develops to allow for the integration of additional senses into interactive experiences, there is a need to bridge the divide between the real and the virtual in a manner that stimulates the five senses consistently and in harmony with the sensory expectations of the user. Applying the philosophy of a neurological condition known as synaesthesia and crossmodal correspondences, defined as the coupling of the senses, can provide numerous cognitive benefits and offers an insight into which senses are most likely to be ‘bound’ together. This thesis aims to present a design paradigm called ‘virtual synaesthesia’ the goal of the paradigm is to make multisensory experiences more human-orientated by considering how the brain combines senses in both the general population (crossmodal correspondences) and within a select few individuals (natural synaesthesia). Towards this aim, a literature review is conducted covering the related areas of research umbrellaed by the concept of ‘virtual synaesthesia’. Its research areas are natural synaesthesia, crossmodal correspondences, multisensory experiences, and sensory substitution/augmentation. This thesis examines augmenting interactive and multisensory experiences with strong (natural synaesthesia) and weak (crossmodal correspondences) synaesthesia. This thesis answers the following research questions: Is it possible to replicate the underlying cognitive benefits of odour-vision synaesthesia? Do people have consistent correspondences between olfaction and an aggregate of different sensory modalities? What is the nature and origin of these correspondences? And Is it possible to predict the crossmodal correspondences attributed to odours? The benefits of augmenting a human-machine interface using an artificial form of odour-vision synaesthesia are explored to answer these questions. This concept is exemplified by transforming odours transduced using a custom-made electronic nose and transforming an odour's ‘chemical footprint’ into a 2D abstract shape representing the current odour. Electronic noses can transform odours in the vapour phase generating a series of electrical signals that represent the current odour source. Weak synaesthesia (crossmodal correspondences) is then investigated to determine if people have consistent correspondences between odours and the angularity of shapes, the smoothness of texture, perceived pleasantness, pitch, musical, and emotional dimensions. Following on from this research, the nature and origin of these correspondences were explored using the underlying hedonic (values relating to pleasantness), semantic (knowledge of the identity of the odour) and physicochemical (the physical and chemical characteristics of the odour) dependencies. The final research chapter investigates the possibility of removing the bottleneck of conducting extensive human trials by determining what the crossmodal correspondences towards specific odours are by developing machine learning models to predict the crossmodal perception of odours using their underlying physicochemical features. The work presented in this thesis provides some insight and evidence of the benefits of incorporating the concept ‘virtual synaesthesia’ into human-machine interfaces and research into the methodology embodied by ‘virtual synaesthesia’, namely crossmodal correspondences. Overall, the work presented in this thesis shows potential for augmenting multisensory experiences with more refined capabilities leading to more enriched experiences, better designs, and a more intuitive way to convey information crossmodally.
| Item Type: | Thesis (PhD) |
|---|---|
| Divisions: | Faculty of Science and Engineering > IDEAS |
| Depositing User: | Symplectic Admin |
| Date Deposited: | 20 Jul 2023 14:44 |
| Last Modified: | 20 Jul 2023 14:44 |
| DOI: | 10.17638/03168935 |
| Supervisors: |
|
| URI: | https://livrepository.liverpool.ac.uk/id/eprint/3168935 |
Dimensions
Dimensions
