Standard input devices for mobile phones are directional keys and discrete thumb-joysticks. These devices are dedicated to the discrete GUIs of the phones (eg. scroll lists and small icons arrays). Today, new mobile applications are arising and require adapted interfaces. In particular, the widespread of 3D applications will be favored if users can efficiently point on any part of thescreen. In this paper, we propose a new interaction technique called Jump and Refine for selection tasks on mobile phones. This technique is based on two levels of cursor displacement in order to reduce the number of keystrokes. The first level allows fast movements into an underlying grid. The second one can be used for accurate positioning into the selected area. We present a user study which shows that using a first coarse jump level decreases the selection completion times. The study also shows that the technique is widely accepted by the users. Finally, we discuss the optimal grid sizes.

This paper presents some experimental results on the comparison of users performance for different kinds of 3D interaction tasks (travel, manipulation), when using either a standard desktop display or a large immersive display. The main results of our experimentation are the following: first, not all users benefit similarly from the use of large displays, and second, the gains of performance strongly depend on the nature of the interaction task. To explain these results, we borrow some tools from cognitive science in order to identify one cognitive factor (visual attention) that is involved in the difference of performance that can be observed.

We present the CAT (Control Action Table), a 6 degrees of freedom freestanding input device designed for interaction with Virtual Environments displayed on huge screens. Both isotonic and isometric sensing modes allow the user to easily and efficiently perform 3D interaction techniques. A 2D tablet fixed on the tabletop allows them to perform accurate 2D interaction techniques.