Barbara Dosher¹ (), Songmei Han¹, Zhong-Lin Lu²; ¹University of California Irvine, ²NYU-Shanghai & New York University

Object attention refers to preferred processing of multiple aspects of one object rather than across objects. We examine the mechanisms of object attention by manipulating the contrast of external noise added to the stimulus. We applied the perceptual template model (PTM) to evaluate external noise exclusion and stimulus enhancement (internal additive noise reduction). We compared subject’s performance in reporting the orientation and phase of one peripheral Gabor with their performance in reporting the orientation of one Gabor and the phase of another (single versus dual object report). Dual object report deficits are the classic demonstration of object attention. Two Gabors were presented in each trial (±7.3 deg eccentricity on the horizontal meridian; vertical +/-10 deg orientation; sine or cosine phase; Gabor window sigma = 0.45 deg). Single and dual object report were blocked, and the location of first (or sole) object report was indicated by a precue. External noise contrasts were [0, 0.02, 0.04, 0.08, 0.17, 0.33], and each external noise was tested at 7 stimulus contrasts, yielding 24 psychometric functions (6 external noises x 2 judgments x 2 attention conditions) in two-alternative judgments (n= 5 observers). The data were interpolated at 65%, 75%, and 85% correct to generate threshold versus external noise contrast (TvC) curves, which were fitted (average r2=.9812) with an extended PTM (ePTM) to account for discrimination between overlapping (non-orthogonal) target templates. The single-object report conditions showed mixtures of external noise filtering and stimulus enhancement (reduced external noise and internal additive noise with Af= 0.76 and Aa= 0.62 for orientation judgments; Af= 0.87 and Aa=0.88 for phase judgments) and more narrowly tuned perceptual templates relative to dual-object report. Distributing attention across two objects for different judgments (orientation and phase) is mediated by less precise perceptual templates.

Acknowledgements: Supported by NEI (# EY–17491).