| Experimental studies on object recognition generally use static objects as stimuli. In the real world, however, motion abounds---the observer and object often move around in the environment. Research in this thesis explores the effects of motion and temporal factors on the processes involved in object recognition. The first four experiments examine explicit memory, as tested by "old/new" recognition, while the last two experiments examine implicit memory as tested by a "symmetric/asymmetric" object decision task. All experiments employ a study-test block design.;The first two experiments compare the recognition of moving objects with the recognition of stationary objects. Experiment 1 shows that translational motion during study facilitates later recognition, but the same motion during test impairs recognition. Experiment 2 shows that looming/zooming motion during study impairs recognition, while having no effect during test. These results demonstrate that motion information can indeed affect the encoding and retrieval processes in object recognition.;Experiment 3 examines the role of spatiotemporal information in object recognition, using continuously rotating objects as stimuli. It is found that when objects reverse direction of rotation from study to test, recognition performance is impaired. This result suggests that spatiotemporal information can be encoded in object representations.;In Experiment 4, the effect of temporal coherence on recognition is investigated. Image sequences differing in degree of temporal coherence are presented for study. Results show that in general, higher temporal coherence leads to higher recognition. However, a temporal association hypothesis that predicts viewpoint invariant recognition is not supported by the data.;Experiment 5 shows that reversal of rotation direction abolishes priming effects, suggesting that spatiotemporal information can be also encoded in implicit memory representations. However, priming is not influenced by temporal coherence manipulations in Experiment 6, a result possibly caused by a lack of statistical power.;Collectively, the data argue for a consideration of motion and temporal information in object recognition research. The traditional approach of separate processing streams for form and spatial vision should be extended to account for the motion-form interactions demonstrated here. |
