Exploring interaction modes for image retrieval

The number of digital images in use is growing at an increasing rate across a wide array of application domains. That being said, there is an ever-growing need for innovative ways to help endusers gain access to these images quickly and effectively. Moreover, it is becoming increasingly more difficult to manually annotate these images, for example with text labels, to generate useful metadata. One such method for helping users gain access to digital images is content-based image retrieval (CBIR). Practical use of CBIR systems has been limited by several "gaps", including the well-known semantic gap and usability gaps [1]. Innovative designs are needed to bring end users into the loop to bridge these gaps. Our human-centered approaches integrate human perception and multimodal interaction to facilitate more usable and effective image retrieval. Here we show that multi-touch interaction is more usable than gaze based interaction for explicit image region selection.