I second everything Andrew said and would add this; in my experience any
eyetracker or HMD will cause a headache after ~45 minutes if it is
"tight enough." My generic advice to people learning how to set up the
ASL or ISCAN headband-mounted systems is to tighten the band, then
loosen it until the observer says it's comfortable, then back it off
another notch because it will invariably feel too tight after ~15
minutes. In experiments where observers have to move enough to cause
problems with shifting headgear we limit time to ~20 minutes or
preferably use the goggle or eyeglass frame system.
It's worth noting that the "pupil-corneal reflection" system that
virtually all video-based eyetrackers use to compensate for movement of
the eye optics with respect to the head do not compensate correctly for
systems with head-mounted scene cameras. While the CR compensation can
do a good job of nulling shifts of the eye camera on the computed
eye-in-head position, that value is used to superimpose the gaze cursor
over the scene image, which will be aimed in the wrong place if the
headband has slipped. This only applies to the video record - the
correction to the eye-in-head data stream doesn't suffer from this
problem unless the headgear slips during calibration.
It's theoretically possible to extract the headband slippage from the
pupil & CR data, but as far as I know it isn't used in current systems.
Jeff Pelz
Visual Perception Lab
Rochester Institute of Technology
Andrew Duchowski wrote:
>On Mon, May 24, 2004 at 03:04:56PM -0700, Jeff Mulligan wrote:
>
>
>>>I am wondering if anyone has ever encountered problems with subjects getting
>>>headaches after using the Iscan head-mounted eye tracker for more than 45
>>>minutes (or any other problems with experiments of this length).
>>>
>>>
>>I have an ISCAN head-mounted system that was purchased in the mid-90's,
>>[..]
>>
>>
>
>Similar setup here, but binocular (VR V8 HMD). Our main problem is
>the heat. The bands on the helmet would indeed pose a problem if too
>tight. Our secondary problem is that for the optics to properly image
>the eyes, one has to wear the helmet rather low. This causes problems
>by riding on the nose for some users. We've had ISCAN alleviate the
>problem by filing down the square brackets and that's helped, but it
>can still touch the bridge of the nose which isn't all that comfortable.
>
>But the heat is the main problem. Each eye imaging apparatus has two
>IR LEDs, and these tend to heat up. Depending on how the user breathes,
>it is possible to fog up the lenses (if one blows out the nose the
>air can go up the lens setms).
>
>
>
>>We have addressed the problem by going to a lighter, more comfortable
>>headmount. We have a unit based on a pair of raquetball goggles, which is
>>essentially a copy of one built at RIT by Jason Babcock. Jason's new
>>design is based on a pair of safety glasses and has eliminated the hot
>>mirror, and looks very nice.
>>
>>I am not familiar with ISCAN's current offerings, but I have no doubt that
>>they have improved during the time since we purchased our unit.
>>
>>
>
>I would second both points although I haven't seen ISCAN's current
>setup (they were distributed by Polhemus as recently as 2002, not
>sure whether they even still sell eye trackers directly or not).
>But yes, the RIT system looks quite nice - their recent ETRA paper
>provides a list of the majority of the part numbers. The paper is
>available via the ACM's Digital Library <htp://www.acm.org/dl>, if
>anyone's interested (I don't have a copy so please don't email me
>about it, sorry).
>
>Cheers,
>Andrew
>
>--
>Andrew T. Duchowski, Assoc.Prof. 451 Edwards Hall (area code: 864)
>Computer Science Dept., MS-0974 tel:656.7677 fax:656.0145 cel:247.2050
>Clemson University mailto:[log in to unmask]
>Clemson, SC 29634-0974 http://www.vr.clemson.edu/~andrewd/
>
>--
>EYE-MOVEMENT mailing list ([log in to unmask])
>N.B. Replies are sent to the list, not the sender
>To unsubscribe, etc. see http://www.jiscmail.ac.uk/files/eye-movement/introduction.html
>Other queries to list owner at [log in to unmask]
>
>
--
EYE-MOVEMENT mailing list ([log in to unmask])
N.B. Replies are sent to the list, not the sender
To unsubscribe, etc. see http://www.jiscmail.ac.uk/files/eye-movement/introduction.html
Other queries to list owner at [log in to unmask]
|