No interaction of first- and second-order signals in the extraction of global-motion and optic-flow

• Published in: Vision research (Oxford) 2011-02, Vol.51 (3), p.352-361
• Main Authors: Cassanello, Carlos R., Edwards, Mark, Badcock, David R., Nishida, Shin’ya
• Format: Article
• Language: English
• Subjects: Biological and medical sciences; Contrast Sensitivity - physiology; Eye and associated structures. Visual pathways and centers. Vision; Fundamental and applied biological sciences. Psychology; Global-motion extraction; Humans; Lighting; Masking; Motion integration; Motion Perception - physiology; Pattern Recognition, Physiological - physiology; Perceptual Masking - physiology; Photic Stimulation - methods; Second-order motion; Sensory Thresholds - physiology; Vertebrates: nervous system and sense organs
• ISSN: 0042-6989; 1878-5646
• DOI: 10.1016/j.visres.2010.11.012

► First-order noise does not mask global-motion or optic-flow extraction from second-order signals. ► Second-order noise does not mask global-motion or optic-flow extraction from first-order signals. ► Evidence supporting independent first- and second-order global-motion systems. Edwards and Badcock ( Vision Research 35, 2589, 1995) argued for independent first-order (FO) and second-order (SO) motion systems up to and including the global-motion level. That study used luminance (which they called FO) and contrast (SO) modulated dots. They found that SO noise dots did not mask signal extraction with luminance increment dots while luminance increment dots did mask SO signal extraction. However, they argued this asymmetry was not due to a combined FO–SO pathway, but rather due to the fact that the luminance-modulated dots, being also local variations in contrast, are both FO and SO stimuli. We test their claim of FO and SO independence by using a stimulus that can generate pure FO and SO signals, specifically one consisting of multiple Gabors (the global-Gabor stimulus) in which the Gaussian envelopes are static and the carriers drift. The carrier can either be luminance-modulated (FO) or contrast-modulated (SO) and motion signals from the randomly-oriented local Gabors must be combined to detect the global-motion vector. Results show no cross-masking of FO and SO signals, thus supporting the hypothesis of independent FO and SO systems up to and including the level extracting optic-flow.