Abstract
Human observers can perceive their direction of heading with a precision of about a degree. Several computational models of the processes underpinning the perception of heading have been proposed. In the present study we set out to assess which of four candidate models best captured human performance; the four models we selected reflected key differences in terms of approach and methods to modelling optic flow processing to recover movement parameters. We first generated a performance profile for human observers by measuring how performance changed as we systematically manipulated both the quantity (number of dots in the stimulus per frame) and quality (amount of 2D directional noise) of the flow field information. We then generated comparable performance profiles for the four candidate models. Models varied markedly in terms of both their performance and similarity to human data. To formally assess the match between the models and human performance we regressed the output of each of the four models against human performance data. We were able to rule out two models that produced very different performance profiles to human observers. The remaining two shared some similarities with human performance profiles in terms of the magnitude and pattern of thresholds. However none of the models tested could capture all aspect of the human data.
Highlights
Optic flow is the pattern of optical motion available at the eye during relative movement between the observer and the scene
For LHP80 thresholds do appear to depend on the quantity of flow when the stimulus contains no noise this dependence is markedly reduced when noise is added to the stimulus
We note that WC99 appears to differ from the other models in the sense that performance continues to improve beyond the 100 dots per frame level
Summary
Optic flow is the pattern of optical motion available at the eye during relative movement between the observer and the scene. It is known that primates are sensitive to the stereotypical patterns of optic flow which arise when an observer moves through a largely stationary scene (Figure 1). An additional role proposed suggests that that optic flow drives rapid eye movements which act to stabilize the foveal image and maintain correspondence between images on the two retinae (e.g., see Busettini et al, 1997). This optic flowdriven stabilization process helps to preserve foveal visual acuity and stereo vision during observer movement (Angelaki and Hess, 2005). In the flow parsing hypothesis, it has been suggested that optic flow processing plays an important role in the assessment of scene-relative object movement during self movement (Rushton and Warren, 2005; Rushton et al, 2007; Warren and Rushton, 2007, 2008, 2009a,b; Matsumiya and Ando, 2009; Pauwels et al, 2010; Warren et al, 2012)
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
