A VR-Based Hybrid BCI Using SSVEP and Gesture Input

Hands-free interaction in immersive virtual reality: A systematic review

“…Similarly to eye interaction, one can obtain the head gaze (i.e., the direction of the head forward vector), and by overlaying a reticle or point drawn in the middle of the image, users can point at objects with their heads, enabling symbolic input by pointing the head to a keyboard (e.g., [42]), and most often, to perform selection tasks (e.g., [114,115]). Less commonly, it is also possible to detect head gestures by using its relative movements, which can be used to, for instance, instruct the system to reset the current action [37]. Interaction with head gaze is useful for minimal VR setups since it does not require additional hardware.…”

“…With steady-state visually evoked potentials (SSVEP), it is possible to read a signal response on the brain by providing a visual stimulation at a certain frequency (i.e., elements flashing at a specific rate). This can be used, to determine if the user is focusing its visual attention on a particular key in a virtual keyboard [37,68], allowing symbolic input. Besides SSVEP, Event-related potentials (ERP) also account for other types of stimulation, such as the P300 wave which denotes a reaction to a stimulus.…”

“…The second is to use a virtual keyboard in the VE and use the selection methods (e.g., eye gaze [95] and head gaze [42]) to choose which letters to input. In [37] a novel approach was used, with a keyboard similar to a T9 keyboard consisting of three main keys that when selected would consecutively show a subset of the previous letters available until only one letter was selected, being the keys selected by a BCI with SSVEP. In [68] a combination of eye-tracking and BCI SSVEP was used to accurately select the keys.…”

“…The satisfaction component is measured with user questionnaires. Most of the studies ask users on their sentiment towards the interfaces being evaluated and their feedback on them (e.g., [37,114,115,119]) and which interfaces they preferred to use (e.g., [31,117]) through custom made questionnaires, some of which (e.g., [2,85]) based on ISO 9241 [49] . A few validated questionnaires for user sentiment and preference were the User Experience Questionnaire [64] used by [98], the Positive and Negative Affect Schedule Survey [19] used by [71,72], and the AttrackDiff Questionnaire [44] used in [25].…”

“…Regarding efficiency, the most used evaluation metric is the time users take to complete tasks using the interfaces (e.g., [8,35,98]) and respective speed [114,115]. Performance through the number of tasks users can accomplish during a fixed period is also a preferred evaluation metric (e.g., [75,113]), especially for symbolic input (e.g., [84]), which is also evaluated using the Information Transfer Rate ( [37,68]). Somewhat less used is Fitt's Law [28] to calculate the performance throughput of selection tasks ( [85,119,120]).…”