Hastening Orientation Sensitivity

1
Hastening Orientation Sensitivity
262
262
Nestor Matthews, Kei Kurosawa Kristen Strong
Department of Psychology, Denison University,
Granville OH 43023 USA
Results
Discussion
Introduction
Training significantly hastened fine orientation
sensitivity temporal resolution improved by 51
at the cardinal axis, and 86 at the oblique
axis. The training also sharpened angular
resolution significantly at each axis. This
sharpening was specific to the trained axis for
each group. Why did cardinal learning occur?
Here are two possibilities. Perceptual learning
on a variety of visual tasks tends to be greater
when external noise is added3. Studies using
external noise4,5 suggest greater internal noise
along oblique axes, which may explain why greater
learning occurs at oblique axes. Here, rather
than adding external noise, we exploited the
imprecision that occurs naturally at extremely
brief durations6,7. An alternate explanation for
the present data is that the participants learned
to ignore specific orientation bands within the
mask. Mask-inhibition effects have been observed
previuosly8 and would effectively extend the
stimulus duration, thereby generating greater
precision7.
Previous studies1,2 showed that the ability to
see subtle angular differences can improve with
practice at oblique but not cardinal axes. The
cause of this anisotropy is uncertain, and it is
not known whether the same anisotropy pertains to
temporal resolution the briefest stimulus
duration needed to achieve a specified angular
resolution. Here, we used visual masks to
investigate the extent to which fine orientation
sensitivity could be both hastened and sharpened,
at cardinal and oblique axes.
Pre-training
Hastening
Method
Pre / Post Training- The stimuli were
sequentially presented gratings that were
preceded and followed by a bulls-eye mask (as
shown below). Participants judged whether the
second grating in each pair was clockwise or
anti-clockwise to the first. The stimulus
duration across all 1,000 trials was 200 msec,
and the angular difference ranged from -12 to 12
degrees. Cardinal-axis (horizontal) trials were
blocked separately from oblique-axis (diagonal)
trails. At each axis, 84 angular thresholds were
computed for each participant. Participants were
then split into two statistically
indistinguishable groups for training.
Training- During training each participant
practiced at only one axis. Stimuli were
presented at varying durations (8-158 msec) and
the angular difference was equal to twice the
pre-training 84 threshold. There were 10 blocks
of 70 trials in each daily session. Each
participant completed 5 days of training.
Sharpening
Specificity
The Bottom Line
The visual systems response to subtle angular
differences can be both hastened and sharpened,
at cardinal and oblique axes. Sufficient initial
levels of neural imprecision by be necessary for
perceptual learning to occur.
References

1. Vogels Orban (1985) PMID 3832592.
2. Matthews Welch (1997) PMID 9038408.
3. Fine Jacobs (2002) PMID 12678592.
4. Heeley et al. (1997) PMID 9068823.
5. Dakin, Mareschal Bex (2005) PMID 16441191.
6. Ringach, Hawken Shapley (1997) PMID 9153392.
7. Matthews, Rojewski Cox (2005) PMID 15929646.
8. Karni Sagi (1991) PMID 2052578.

This work can be viewed and downloaded
at http//www.denison.edu/matthewsn/vss2006.html
and the Journal of Vision (in press)