Nonhomogeneous compensation for smooth pursuit

1
Non-homogeneous compensation for smooth pursuit

• Camille Morvan Mark Wexler
• LPP(A)/CNRS
• Collège de France Université Paris V

2
s
r
r
3
e
P
s
r
4
k e
P
s
r

• P r k e
• ? P// r// ke
• ? P? r?

5
Procedure
Vy (/s) 5 10 20 40
Vx (/s) /- 60, 30, 20, 10, 0
Target 20/s

• Stim directions from 5 to 175, speeds 5 to 72
  /s
• Stim duration 100 ms
• Stim radius 0.2
• Task report perceived direction
• 5 subjects, 2520 trials

6
Results
s
r
s
r
7
Results

• ? should be a constant. Here it depends on the
  horizontal speed of the stimulus, why?
• The eye velocity estimation depends on retinal
  speed
• Minimization of the perceived speed

8
Minimization hypothesis

• P// m (r// k e)
• Evaluation of m across all conditions (multiple
  regression)
• m 0.72 (R2 0.74)

9

• Evaluation of k ? P// /m r// k e

k 1.32 (0.24)
k 0.91 (0.25)
k 0.51 (0.24)
k 0.28 (0.23)
Ret vel mean -55
Ret vel mean -22
Ret vel mean 31
Ret vel mean 0.5
10
Complete model

• P// mr// (a br//)e
• 0.83 (0.03)
• 0.31
• b -0.010

11
Discussion

• Why would k depend on the retinal speed?
• To get a better evaluation of e
• Why would the visual system minimize the
  perceived speed?
• In other situations of motion perception
  (aperture problem, apparent motion) ambiguity is
  solved by the system by choosing the minimum
  speed.
• Use of this study.

12
Thank you