This is an old revision of the document!
https://www.dokuwiki.org/plugin:latexport#latexport_plugin
to be exported into latex
Use of augmented reality for cognitive psychology studies
Augmented Reality for cognitive psychology and forensic sciences studies
Abstract
The availability in 2021, on the pro-consumer market of standalone (un-tethered) Head Mounted Displays (HMDs) for Virtual Reality (VR) and Augmented Reality (AR) is an enabling technology in the field of cognitive psychology of the human visual system. It is of special interest in forensic sciences where one need to reconstruct the characteristics of an observed phenomena from a witness testimony.
This article tries to list the known issues to which investigators have been confronted ; the new issues raised and which parameters are now measurable. We also list the expected practical limitations of the current state of the art.
Such HMDs allow for much easier measurement of some characteristics of the human visual systems. Some issues existing in more previous methods are circumvented, new issues arise.
A practical setup is described.
Introduction
Parameters Considered
We are here specifically interested in the estimation of
- distance
- absolute size
- angular size
- color
- apparent luminosity / surface brightness
- azimut
- angular height
and the estimation of the evolution of all the parameters over time (includes angular speed)
Known issues
We already know that some parameters are very often highly mis estimated by witnesses.
These problems have been noticed more than a century ago 1), yet, we feel that there is currently a huge lack of characterization of the biaises in the reports.
Some tentative studies were done, but with a lack of control and strict definition of the setup 2).
Some empiric methods currently used may have hidden biaises.
Because one cannot almost never reproduce the real stimuli in the real environment, investigators use basic methods that are at a risk of introducing biaises.
A portion of the population have not assimilated the notion of angular size. It is a source of mis-understandings.
If an optical system with partial obstruction of the view is used, there is a risk of altering the response.
Even though a comparison with a close by object 3), seems to be a good way to go, there is still a risk of altering the response because it implies monocular comparison.
Many factors alter the estimations
Alteration of the estimation due to the context of the observation,
- by looking through a window, depending on the distance to the window
- more generally, by the visual context around the object observed
- by being sitting or standing.
- by the location of the witness : inside, outside of a building.
- by night or day
Alteration by Human factors
- by the emotions felt.
- effect on the estimation of the duration
- potential effect on how the surrounding sound is recalled
Alteration by the apparatus used for simulation
- by looking though an eye piece
- by the use of an object held by hand for comparison (vergence conflict).
This can alter the estimations of
- distance
- absolute size
- angular size
- color
- apparent luminosity / brightness
- azimut
- angular height
? angular speed
sounds
New interesting development : being able to obtain error bars as a function of the witness estimations.
the literature shows some interests for therapeutic use military applications are concentrated on the
google “augmented reality” “cognitive psychology” +“forensic sciences” human factors
State of the art
In 2021, on the pro-consumer market, the HoloLens 2 HMD (figure \ref{fig:hl2}) is considered as the state of the art.
It is of the Optical See Through (OST) type : the virtual objects are “reflected” 4) on a transparent visor. The virtual objects appear somewhat like semi transparent holograms.
Picture
On purpose, for better contrast of the virtual objects over the real world, the visor is tinted (neutral grey). Its transmittance is about 40%.
- Type : AR, optical see through (OST)
- Visualization through a tinted visor
- Field of view : 43° horizontal x 29° vertical. 52° diagonal.
- Resolution : 1440 x 936 5). If these are square pixels, makes an aspect ratio of 1.538 and matches the field of view aspect ratio .
- Angular Resolution : should be 31.58 pix/° at the center of the image and 36.58 pix/° at the edge.
- Display Brightness : 500 nits (candela/m²). 6)
- Virtual Image projected at 2 meters 7)
- Display Type : lasers scanning + waveguides
- Hand Tracking : Yes
- Eye Tracking : Yes
- Stereo Audio