Introduction
Background of head mounted display
The concept of head-mounted displays has already appeared in the consumer market, but never before. Furthermore, other products, including the Oculus Rift [11], head-mounted displays have attracted the attention of large audiences and developers. Oculus Rift is a lightweight headset that allows a user to engage in virtual reality-based interactions.
Problem statement and scope
Multi-finger interaction gripping can provide a much better, firm grip in various one-handed/two-handed poses, allowing the user to rotate the phone or type on the keyboard with various tablet/phone layout adjustments [ 6 , 7 ]. In InfiniTouch [35] they also used capacitive sensing, where the sensors are placed on the front, back and three sides of a phone, allowing the users to touch the entire device. In Pre-Touch [25] it is possible to detect the user's capacitive touch on the top of the screen.
Research contribution…
Interaction on the front side of the spectacle lens has been studied [36, 67], but the top and bottom side-based interaction has not yet been studied for the glasses, four sides of the edges surrounding the front screen of watches have been studied, the mobility situation the result of these studies is unknown. Furthermore, we know very little about target selection on the various input surfaces of the touchpad on glasses such as top, front and bottom surfaces. Google Glass is 70.2mm tall with a 10.4mm wide touchpad on the right side of their glass which is only able to sense the touches on the front [1].
Research aim
Since our research goal is to facilitate interaction in the limited input space and which also include non-visual data.
Research Hypothesis
Thesis structure
Overview
Hands-free interaction
On-body interaction
Mid-air interaction
Hand-Worn interaction
Graspable interaction
Around interaction
Region interaction
Mobility
Overview of Hardware and software system
Hardware prototyping and Software Implementation
- Touch sensor mounted right side of the glass
- Briefing of MPR121 sensor board touch capacitance
- Wiring of the MPR121 sensor boards and wireless data transfer protocol …
- Wiring of the MPR121sensorboardson Epson Moverio BT-200 glass
- Touch capacitance specification and demonstration …
- Novelty of the prototype
- Design rationales of the prototype
Whereas, our custom prototype is able to sense the top, front and bottom sides of the touchpad mounted on the side of the glasses. Whereas, our custom prototype is able to sense the top, front and bottom sides of the touchpad mounted on the side of the glasses. In one-dimensional text input [67], they used the front side of Google Glass to input text into the glasses.
To compare the difference between the MTCT of the four conditions, we performed a 2 x 2 two-way repeated measures ANOVA, taking body position (sitting and walking) and temporal recording (sequential one-finger tapping and simultaneous two-finger tapping) as the two independent variables to measure the dependent variable MTCT. In the case of the time demand parameter, the mean time (time pressure) workload requirement for sequential one-finger tapping was higher than for simultaneous two-finger tapping, but body positions did not affect time (time pressure) workload for both sequential and to tap at the same time. . Proceedings of the Supplementary Publication of the 26th Annual ACM Symposium on User Interface Software and Technology, ACM.
Proceedings of the 17th International Conference on Human-Computer Interaction with Mobile Devices and Services, ACM. Proceedings of the 10th International Conference on Human Computer Interaction with Mobile Devices and Services, ACM. Proceedings of the 12th International Conference on Human Computer Interaction with Mobile Devices and Services, ACM.
Proceedings of the 18th International Conference on Human-Computer Interaction with Mobile Devices and Services, ACM.
Experimental Design, Result and Analysis
Briefing of Experimental Design, Result and Analysis
In terms of system or product usability, both quantitative and subjective measurements are an important aspect. Therefore, we also captured the subjective measure of NASA TLX, including how mentally demanding the task was, the physical level of the task, the temporal demand in terms of time pressure that users went through, the user's perception of their performance, accuracy level, effort users had to put in to reach the performance level and what was the level of frustration they experienced in all four different conditions [65].
Motivation …
Experimental Design
- Participants
- Task design
- Experimental task
- Procedure
- Novelty of the study
In their study, for the front and both left and right sides of the touchpad, they used built-in both landOn and liftOff target selection techniques, and in the eye-free condition, it reduced errors by about 40%. Few examples of 36 combinations of a pair of targets for nine locations of top, front and bottom sides are shown in (Figure 6), all the combinations can be found in Appendix 1. To conduct the experiment, we used the 36 combinations and the participants completed each of the 4 conditions (Body positions: sitting, walking; Temporal tapping: single finger sequential tapping, pair finger simultaneous tapping, all conditions were fully counterbalanced by the Latin square design) with 3 blocks of repetitions.
In total, each study participant generated 432 trials, block 1 of each condition was treated as practice trials, so the remaining 288 trials for each participant were retained for the final analysis. The experiment was designed within-subjects so that each of the participants had to complete all four conditions with the appropriate order of the conditions as shown in (Table 3). Then, a fixation point in the form of a circular shape could be seen in the center of the glass of the 10 x 10 pixel screen, which remained on the screen for 500 ms.
The feedback for the correct trial was given as a green colored circle and the incorrect trial was given as a red colored circle, which was shown in the center of the screen for approximately 1000ms. In the middle of the room there was a large meeting table 2.1 meters high and 1.2 meters wide. As well as starting the walking condition study, participants had to complete five laps as before while wearing glasses while walking.
No previous study, especially for glass-based input/target selection, implemented this single-finger sequential and pair-finger simultaneous touch on three sides of the touchpad and it can benefit from both temporal tapping in different posture situations.
Experimental Results
- Quantitative analysis
- Mean Task completion time (MTCT)
- Mean Task Error Rate (MTER)
- Subjective analysis
- Mean Rating of Workload (MRW)
- Descriptive analysis
Temporal touch had 2 levels (sequential one-finger tapping and simultaneous two-finger tapping), and body pose had 2 levels (sitting and walking). Which shows that, in one-finger temporal tapping, the overall sequential error rate was higher for the standing body pose and, in one-finger temporal tapping, the simultaneous tapping error rate was higher high in walking condition. This suggests that neither the sequence of a single finger and simultaneous two-finger tapping, nor the standing or sitting body pose, had a measurable impact on the mental load demand experienced in the study.
This suggests that neither the single-finger sequence and simultaneous two-finger tapping nor body position while walking or sitting had a measurable effect on the exercise demand in the study. Which shows that in single-finger temporal tapping, the sequential temporal (time pressure) workload was higher in the sitting body position, whereas in paired-finger temporal tapping, the simultaneous temporal (time pressure) workload was higher in walking. This suggests that neither the single-finger sequence and simultaneous two-finger tapping nor body position while walking or sitting had a measurable effect on the workload of the study participants.
This suggests that neither the single-finger sequence and paired-finger simultaneous tapping nor the walking or sitting body posture had a measurable impact on the given workload experience of the study participants. Which indicates that sequential experienced frustration in one-finger temporal tapping was higher for the walking body posture, whereas simultaneous experienced frustration was higher in the sitting posture. This suggests that neither the single-finger sequence and paired-finger simultaneous tapping nor the walking or sitting body posture had a measurable impact on the participants' overall workload experienced in the study.
Which shows that in the temporary one-toe strike the overall follow-up load rating was higher for the standing posture, while in the temporary one-toe strike the overall workload rating was higher in the sitting down
Overall Discussion
Conclusion
We also want to mention that, 11 combinations (farthest touch combination with 8 and 3 on the lower side of the touch) of simultaneous touch with pairs of fingers were more error prone and took longer to complete the task . While we facilitated the landOn and liftOff technique for target selection, which allows participants in the situation without eyes. Participants also mentioned that "I think my performance could be worse if there was no such technique of landOn and liftOff input".
Previous research reported that the smartphone-based target selection technique, where the authors only enabled the landOn technique and performance, was more error-prone [ 42 , 53 ]. We want to emphasize that in smartphones/smartwatches, the screen can be seen visually, and in our case, it supports time tapping without eyes on the side-mounted touchpad on wearable glasses. Our study confirms that the two-finger tap-based input was suitable, having a similar success rate of 85% and shorter completion times for both walking and sitting.
Our system utilizes both landOn and liftOff, which enables the user to make the selections successfully and comfortably. We adapted our system in a way that allows users to spontaneously grab each edge, which helped them type accurately as they went. Whereas, single-finger tapping-based input had the highest success rate of 90% in the seated condition among all conditions, indicating that this type of interaction is possible in the seated condition, on the other hand, single-tapping had the lowest success rate of 79%, suggesting that we that these types of interactions are not appropriate while users are engaged.
Limitations and future work
2014 ACM International Joint Conference on Pervasive and Ubiquitous Computing, UbiComp 2014, Association for Computing Machinery, Inc. You'll tap the top, front, and right side of a touch input sensor mounted on the right side of a pair of smart glasses. Equipment you will wear in this study includes a pair of smart glasses equipped with a custom touch sensor and a lightweight backpack containing equipment to operate these devices.
Each trial is also initiated by a tap, followed by instructions to touch two specific regions on the touch sensor. Sit on the chair and place your elbow on the desk so that your right hand can comfortably reach the touch input sensor on the smart glasses. In each trial, two target areas will be shown on the smart glasses in a 3D view.
You can also move your fingers forward and backward along the sides of the device to refine a selection and lift your finger when you come across the right target. While doing this, you should hold your right hand up so that it can comfortably reach the touch input sensor on the smart glasses.
