Abstract
Top-down attention can affect apparent motion (AM) of sequentially presented static tokens (e.g., Wertheimer, 1912; Kohler et al, 2008). It is also likely that the sequential token presentations can drive attention to shift from one location to another. It is thus reasonable to hypothesize that such stimulus-driven attention-shift itself is a driver of AM perception. We tested this hypothesis using a motion adaptation protocol, where we manipulated the adaptation display (adapter) and attention status for 2 minutes and then measured perceived AM directions with a bistable motion quartet (~ 2x2 deg, 150 msec) for 1 min. We found the following results. (1) A large vertical AM adaptation display (4x4 deg, 200 msec frame duration) can bias the perceived AM direction to be more horizontal, and vice versa. This indicates AM adaptation is not location specific, which is a characteristic that is more closely associated with the attention mechanism than motion mechanism. (2) The adaptation effect was reduced when observers performed a RSVP task centrally while ignoring the more peripherally presented adapting AM display (4x4 deg). This suggests attending centrally prevents possible attention movements between tokens and thus prevents AM adaptation. (3) The adaptation effect occurred when the adapter was a continuous motion stimulus (speed: ~20 deg/sec) when presented in conjunction with the RSVP task. This indicates that AM requires attention, but continuous motion perception does not. (4) The adaptation effect occurred when the adapter was non-moving vertical or horizontal flashing bars (~ 4 deg long, duration=200 msec). This suggests that during the adaptation phase, the flashing bars were able to attract attention and to retard the stimulus-driven attention process. Taken together, these results support the hypothesis that stimulus driven attention-shift from the AM motion tokens can drive AM perception.
Meeting abstract presented at VSS 2018