Sensitivity to gaze-contingent contrast increments in naturalistic movies: An exploratory report and model comparison

摘要: Sensitivity to luminance contrast is a prerequisite for all but the simplest visual systems. To examine increment detection performance in way that approximates natural environmental input of human system, we presented increments gaze-contingently within naturalistic video freely viewed by observers. A band-limited was applied local region relative observer's current gaze point, and observer made forced-choice response location target (≈25,000 trials across five observers). We present exploratory analyses showing improved as function magnitude depended on direction eye movements location, timing presentation, spatiotemporal image structure at location. Contrast discrimination can be modeled assuming underlying an accelerating nonlinearity (arising from nonlinear transducer or gain control). implemented one such model examined posterior over parameters, estimated using Markov-chain Monte Carlo methods. The parameters were poorly constrained our data; strong priors taken previous research showed poor cross-validated prediction performance. Atheoretical logistic regression models better provided similar model. Finally, explored properties extended incorporates both movement content features. Models transduction may incorporating data artificial perception settings.