Exploiting Physical Referent Features as Input for Multidimensional Data Selection in Augmented Reality

Embedding data into the physical environment using augmented reality (AR) is a practical approach for data visualization as it offers a large and flexible display space on or around the physical referent, i.e. the physical object to which the data is related. Yet current interaction in such context is often performed using cumbersome dedicated devices, tiring mid-air gestures or awkward on-body input. In this paper, we investigate the use of the physical referent itself as a support for input interaction with an embedded space-time cube (STC) representation. Hence, we first identify the most promising mappings between the physical features of the referent (edges, faces, corners) and the STC dimensions. Then, we design three data selection techniques using the physical referent and compare them to mid-air gestures when performing selection tasks on the STC. Overall, our work demonstrates that using the physical referent to support input interaction with embedded data representations is an efficient and comfortable approach for data selection in standing and sitting situations.