Use this URL to cite or link to this record in EThOS:
Title: High fidelity olfaction simulation for virtual environments
Author: Dhokia, Amar Pravin
ISNI:       0000 0004 7961 1233
Awarding Body: University of Warwick
Current Institution: University of Warwick
Date of Award: 2017
Availability of Full Text:
Access from EThOS:
Access from Institution:
Olfaction is a key human sense due to its close connection to the brain, specifically memory and decision making, along with its effects on behaviour and emotions. Clearly, in order for virtual environments to reach perceptual equivalence with the real world, olfaction needs to be incorporated. Perceptual equivalence is defined as the same cognitive response occurring with the users whether they are exposed to real or synthesized stimuli regardless of the levels of duration, intensity or nature. While there have been previous attempts to use olfactory stimuli in virtual environments, these have been limited and not addressed some key aspects of olfaction and its corresponding stimuli, such the effects of high odour concentrations and smell habituation. This thesis presents a framework for a physically accurate olfactory pipeline in order to help provide high fidelity perceptually equivalent virtual environments. The pipeline consists of stages for capturing, data storage, reproduction, and finally delivering stimuli to end users. In particular this thesis focuses on the final delivery stage. It presents a novel physical calibration for the olfaction delivery system and a new human perceptual calibration procedure to ensure a robust and repeatable experience. For the first half of the delivery stage, this thesis provides the blueprint for a physically accurate smell display. This is subsequently shown be capable of reproducing smell stimuli with both accuracy, to intended real world concentrations, and precision, such that the outputted olfactory stimuli are consistently presented at the specified level. In addition, similar to other existing olfactory displays, this smell display is a low cost solution as well as a straightforward design with the intrinsic ability to provide virtually, immediate temporal displacement. The validation of the physical calibration of the display is based on fundamental laws of chemistry to provide the same result under the same specified conditions. For the second half of the delivery stage, a perceptual calibration procedure is presented to calculate Just Noticeable Differences for olfactory stimuli. The purpose of this is to be able to create normalised stimuli levels which are perceivable by the general population but are also perceptually similar across different odours. The work provides two Just Noticeable Difference stimuli levels for three target single molecule odours which have been calibrated on the olfactory display based on a sample population (N=10). These determined stimuli levels can subsequently be utilised as generalisable points for further experimental use with participants when investigating both olfactory phenomena and relationships between olfactory and other sensual stimuli. A final experimental study is proposed in this thesis to clearly demonstrate the capabilities of the research and to explore the possibility of olfactory attention-masking phenomena, similar to those seen in the visual domain. Participants were asked to preferentially rank the odours presented to them based on images of the smell source. Odours were the same as those used in the perception calibration, and were presented as randomised dual conditions either in combination with the other odours or with a blank. Participants were consistently able to correctly identify the correct odours with the correct images, ranking the correct answers either primary or secondary. In addition, a number of preferences for some odours over others were identified. In conclusion, this thesis outlines a novel framework for addressing the physical and perceptual aspects of olfaction in order to provide an accurate representation of a real world equivalent olfactory experience. Experiments show humans are readily able to distinguish between odours when presented at the same time and the evidence obtained suggests there may be attention-masking phenomena in the olfactory domain.
Supervisor: Not available Sponsor: Jaguar Land Rover ; Engineering and Physical Sciences Research Council
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available
Keywords: QA76 Electronic computers. Computer science. Computer software