December 2001
Volume 1, Issue 3
Vision Sciences Society Annual Meeting Abstract  |   December 2001
Learning motion discrimination without MT
Author Affiliations
  • Zili Liu
    Rutgers University-Newark, Newark, NJ, USA
  • Hongjing Lu
    Rutgers University-Newark, Newark, NJ, USA
  • Ning Qian
    Columbia University, New York, NY, USA
Journal of Vision December 2001, Vol.1, 27. doi:
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Zili Liu, Hongjing Lu, Ning Qian; Learning motion discrimination without MT. Journal of Vision 2001;1(3):27. doi:

      Download citation file:

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

  • Supplements

Purpose: to study visual perceptual learning of motion discrimination using paired dot stimuli that are known to suppress MT activity. Method: In a 2AFC task, two random-dot motion stimuli were presented sequentially, each for 200 ms, with an ISI of 500 ms. Each stimulus consisted of 100 pairs of random dots within a circular aperture, and each pair oscillated in opposite directions across each other with an amplitude of 0.3 degrees of visual angle and a life time of 120 ms. The stimuli were thereby locally balanced everywhere, and hence strongly suppressed MT activity (Qian, Anderson, and Adelson, 1994). In each trial, the motion-axes of the two stimuli differed by 8 degrees. The subject judged if the change of motion-axis from the first to the second stimulus was clockwise or counter-clockwise. No static frames of the stimuli provided any information about the motion-axis. This was achieved by positioning one dot pair next to another pair of the same motion-axis so that the four dots formed a parallelogram. The two pairs were positioned so close to each other that the shortest separation was often between dots in different pairs. Therefore, the proximity-based salient orientation defined by two closest dots was often random rather than along the motion-axis. Consequently, the motion-axis of each stimulus was defined by motion alone. 20% of the dot pairs were noise, oscillating along randomly chosen axes. Results: Substantial learning (d′=1.38 to d′=2.69) was obtained. However, when the task was more difficult (5 degrees of angular difference between the motion-axes), learning was impossible even though the performance was well above chance (d′=1.26, 15 sessions, 400 trials each). This result from 1 subject was replicated by 4 additional subjects. Conclusion: MT may not be critical for learning motion discrimination, but task difficulty is critical.

Liu, Z., Lu, H., Qian, N.(2001). Learning motion discrimination without MT [Abstract]. Journal of Vision, 1( 3): 27, 27a,, doi:10.1167/1.3.27. [CrossRef]
 Supported by NSF IBN-9817979 to NQ & ZL.

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.