Figure 4. Flowcharts for the contrast experiments.

(a) The experimental protocol in Experiments 2.1 and 2.2. Four sessions were completed in both the pre- and post-test day, with two in the exposed condition (black bars) and two in the control condition (gray bars). The session sequence was counter-balanced across the subjects. In the exposed condition, stimuli were presented at two diagonal positions (NW/SE or NE/SW) with one adapting orientation (vertical or horizontal). In the control condition, by contrast, stimuli were presented at the other two quadrants with the orthogonal orientation. The positions and orientation of the adapters remained the same within a session. During the 6 training days, subjects finished two adaptation sessions every day as in the pre- and post-tests but only for the exposed condition (black bars). Meanwhile, they completed another two unadapted sessions (white bars) where they were trained in ramp detection at the control positions with no adapters during the ‘adaptation’ phases. The detailed procedures for each session are depicted in Fig. 5a. (b) The experimental protocol in Experiment 2.3. The procedure was similar to Experiment 2.1, except that subjects were not trained in ramp detection at control locations during the training days. (c) The experimental protocol in Experiment 2.4. The pre- and post-tests were identical to the ramp detection paradigm in Experiment 2.2. However, a top-up paradigm was used for training. Subjects finished two adapted (at the exposed locations) and two unadapted (at the control locations) sessions every day. The detailed procedures for each session are depicted in Fig. 6a.