Steven Dakin^1,2 (), Samuel Schwarzkopf^1,2, Selassie Tagoh¹; ¹University of Auckland, ²University College London

Motion adaptation can transiently improve visual acuity (VA; Lages et al, Psych. Sci, v28, p1180-8) which is surprising given that adaptation typically reduces contrast sensitivity (CS). Here we consider the role of information at different spatial frequencies (SFs) in this phenomenon. A single interval 4AFC identification task and a 2IFC contrast discrimination task were used to measure VA and CS respectively. We tested 19 normally sighted observers. VA was measured using both conventional and vanishing tumbling T-optotypes (which contain little low SF structure). CS was measured using SF band-pass filtered T’s (at 3.75, 7.5 and 15 c/deg) presented either in isolation, or on a fixed-contrast SF low-pass “pedestal”. We report that adaptation to receding motion improves VA (by -0.051 logMAR change, p < 0.001) for conventional but not vanishing optotypes (-0.004 logMAR change, p = 0.73), indicating a crucial role for low SFs. Adaptation reduces CS particularly at the lowest SF (by 31%, p < 0.00003). In unadapted observers however, CS for higher SF information was enhanced/facilitated by the presence of low SFs (by 192%, p < 0.00001). We conclude that low SFs are necessary for adaptation-driven acuity enhancement and adaptation reduces sensitivity to them. However since low SF structure facilitates detection of higher SFs, adaptation cannot be improving acuity indirectly by improving detection. Instead suppression of low SFs must directly improve performance at the recognition stage and (it follows) recognition must operate on information that has been combined across SF (cf MIRAGE, Watt & Morgan, 1985, Vis Res, v23 pp1465-77). We show how a MIRAGE-like model - operating on visual information that has been subject to suppression of low SF structure - predicts improved localization of features within letters that could in turn support superior recognition near the acuity limit.

Acknowledgements: Supported by the Marsden Fund Council from Government funding, managed by Royal Society Te Apārangi