A comparison of static and moving presentation modes for image collections

In both professional and personal contexts, a common activity is the search for a target image among a collection of images. The presentation of that collection to a user can assume a wide variety of forms, and it would help interaction designers to be aware of the comparative properties of available presentation modes. A property of major interest is the percentage of correct identification of the presence or absence of the target image within the collection; another is users' acceptance of a presentation mode. Several modes of Rapid Serial Visual Presentation (RSVP) are compared for effectiveness in a number of image identification tasks, and with regard to user acceptance and stated preference.Presentation modes have been classified as static or moving. For a selected representative group of three static and three moving modes, for three image presentation times and for three tasks of increasing complexity, we report experimental results which in most cases establish, with a high degree of statistical confidence, that -- over the range of independent variables investigated - (a) static modes are more successful with regard to identification success than moving modes; (b) static modes are far more preferred than moving ones; (c) identification success generally increases with increase in presentation time per image; (d) for mixed and tile modes, identification success is relatively insensitive to image presentation time; and (e) success rate decreases with increase in task complexity except, notably, for slide-show and mixed modes. Evidence from eye-gaze records suggests that the eye-gaze strategy adopted by a subject exerts a very strong influence on both identification success and mode preference. Conclusions are drawn about guidance that can be offered to an interaction designer.