August 2014
Volume 14, Issue 10
Vision Sciences Society Annual Meeting Abstract  |   August 2014
Perceptual learning of detection of band-limited noise patterns
Author Affiliations
  • Zahra Hussain
    University of Nottingham
  • Patrick Bennett
    McMaster University
Journal of Vision August 2014, Vol.14, 1157. doi:
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Zahra Hussain, Patrick Bennett; Perceptual learning of detection of band-limited noise patterns. Journal of Vision 2014;14(10):1157.

      Download citation file:

      © ARVO (1962-2015); The Authors (2016-present)

  • Supplements

Perceptual learning has most frequently been studied for discrimination and identification tasks, in which learning is specific to stimuli exposed throughout practice. Here, we asked whether practice improves detection of textures in noise, and whether the improvements, if any, are stimulus specific when stimulus features are not easily identified. We used two external noise levels, following previous studies showing different patterns of improvement in discrimination in low and high noise. Two groups of observers practiced detection of five noise textures (2-4 cpi) on two consecutive days. The texture was presented at one of eight contrasts, including a zero-contrast (signal absent) condition. The observer's task was to detect whether the texture was present or absent on each trial (yes/no). One group practiced the task with the same five textures on both days, and the other group switched to five novel textures on Day 2. Noise levels were blocked, and the five textures were randomly presented throughout the session. We calculated d' at each contrast, and the false alarm rate for each observer in each noise level on both days. Performance improved for both groups in high noise, but the improvement was larger for the same texture group, particularly at high contrasts. Performance improved slightly in low noise for the novel texture group, but not for the same texture group. The improvement in low noise was significantly associated with a reduction in false alarms. The results suggest that learning of texture detection generalizes to novel textures from the same bandwidth, but there is an advantage in high noise for previously exposed textures (i.e., stimulus specificity) despite absence of identification. The low noise data are consistent with work suggesting that threshold improvements in certain yes-no tasks are accompanied by changes in the decision criterion.

Meeting abstract presented at VSS 2014


This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.