Edinburgh Napier University

  Human perception is partially affected by what has been previously experienced. These history effects presumably help tackle current sensory uncertainty by tracking past stimulus statistics. However, there is no definitive framework on how stimulus history affects perception at different levels of uncertainty. We asked observers to discriminate the orientation of ambiguous Gabor patches at high or low contrast, while we dynamically changed the orientation statistics of unambiguous high-contrast stimuli. We found both repulsive and attractive history effects at different timescales and differences between high- and low-contrast test patches. We present a computational model that can account for these different history effects by tracking both the volatility of past stimulus statistics and the observer’s uncertainty on the current stimulus. This model may help resolve some conflicting results of history effects in the literature.