Virtual Table Tennis 3d
Assunta Gergely
A few people have started offering in game virtual reality table tennis coaching. I think this is incredible. The coach and player play at the table simultaneously. They can stand on the same side of the table or opposite each other. They can use a ball machine together or hit with one another. They can discuss table tennis and do exercises together all in game.
Suppose the game developers were to add features to improve the coaching experience. What features would be essential? (1) Record and replay hits from a game or practice (2) Track stats during practice (3) Coach multiple people simultaneously across multiple tables (4) Mix and match replays
There are two reasons to want to record and replay hits from a game or practice. One is to learn to defend against them. Say you play against an opponent with a powerful serve or forehand smash. By recording and replaying their attack you can practice defending against it.
A key assumption of VR training is that the learned skills and experiences transfer to the real world. Yet, in certain application areas, such as VR sports training, the research testing this assumption is sparse.
Real-world table tennis performance was assessed using a mixed-model analysis of variance. The analysis comprised a between-subjects (VR training group vs control group) and a within-subjects (pre- and post-training) factor.
Copyright: 2019 Michalski et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work is funded by an NHMRC Dementia Research Leadership Fellowship (GNT1136269) to TL; AS is supported by the University of South Australia Research Themes Investment Scheme and DS is supported by an Australian Government Research Training Program Scholarship. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Using Virtual Reality (VR) as a tool for training is becoming increasingly popular. VR is an immersive digital space where users can interact with objects and navigate as if they are present in a real environment [1, 2]. VR technology has garnered interest for use in situations where training in the real world is logistically difficult to organise, dangerous or impractical [2]. For example, training programs for surgeons [3], pilots [4] and firefighters [5] are taking advantage of the realism and flexibility that VR offers. A key assumption of VR training is that the learned skills and experiences transfer to the real world [6]. Yet, in certain application areas, such as VR sports training, there has been little research testing this assumption.
Applying VR to sports training has several advantages. Firstly, it offers the possibility for people to train without needing access to the necessary sporting environment (e.g., a downhill slope for skiing) or multiple training partners (e.g. football). Secondly, actively incorporating VR into sports training allows users to log their performance and closely monitor their development. Thirdly, VR is scalable and provides a great degree of freedom to create and control virtual environments in diverse ways. These advantages provide an immense opportunity to incorporate evidence-based practices to achieve greater performance gains from training. For example, evidence-based research on training, such as the challenge point framework, suggests that individuals develop their skills best when they are constantly challenged [7]. Integrating this research into a virtual environment can easily be accomplished by adding variability into training and systematically adjusting the level of difficulty based on the ability of the user [8].
There are many examples of training in VR which has led to positive real-world outcomes. For example, when medical students used a medical VR training simulator to learn gallbladder dissection, they were able to perform the task 29% faster in the real world than non-VR trained students and with six times fewer errors [3]. Bliss et al. [9] found that VR could be used to train firefighters equally as well on real world spatial navigation as traditional methods, and Carlson et al. [10] found that VR training improved performance on a real world assembly task. However, researchers such as Kozak et al [11] found that there was no transfer of training from VR to the real world, in this case using a pick and place task. These conflicting results show that there is still more research that needs to be done on transfer of training from VR to the real world.
Despite these positive results, for VR to become an accepted tool in sports training, it must be determined whether it can have a real-world impact. The present study aimed to investigate the transfer effects from VR training to the real world, using the fast-paced sport of table tennis. It was hypothesised that VR table tennis training will lead to significant improvements (pre-test to post-test) in real-world performance (serving and rallying) on both quantitative aspects (serving accuracy and number of rallies without errors) and overall aspects of skill quality (ball height, strength, consistency, technique and coordination) as compared to no training.
No participant had an injury, disability or any other reason which could affect their involvement in the study. To ensure that participants had normal or corrected-to-normal visual acuity their vision was assessed on various tests including: Snellen Eye Chart, RAF Rule and Fonda-Anderson Reading Chart. In addition, participants in the VR training group had normal or corrected-to-normal stereo acuity, as determined by the Butterfly Stereo Acuity test (Vision Assessment Corporation, 2007). None of the participants were competitive table tennis players.
Participants were recruited via flyers which were placed around the University of South Australia (UniSA) Magill Campus. All participants provided informed consent prior to participation and received an honorarium of $20/hour for participation. The study was granted ethics approval from the UniSA Human Research Ethics Committee.
The appropriate sample size was calculated using the G*Power3 software [19]. Experiment 1 in the study by Todorov et al. [13], examined transfer of table tennis skills in a virtual environment via number of target hits before and after virtual training, finding an effect size of 0.86. Using this effect size as an estimate for the Power Analysis, it was calculated that 18 participants per condition would be needed to suffice power (0.80) with α = 0.05.
The HTC Vive head-mounted display (HTC, with technology by Valve Corporation, April 2016) was used. Immersive stereoscopic head-mounted displays (HMD) such as the Vive create a sense of presence in the user by viewing a 360-degree virtual environment that moves in real time in accordance with the movements of the participant [20]. Users were required to wear the HMD while holding two controllers to interact within the virtual environment, one as a device to simulate ball dispersion and another as a simulated table tennis bat. The HMD accommodates users with visual impairments, by allowing them to wear glasses and contact lenses while using the device. The VR apparatus was used in a 1930mm x 3300mm room with two base stations installed in opposite corners, enabling room-scale tracking.
Eleven: Table Tennis VR (developed by Fun Labs) was used. This is a game which requires users to interact by moving and responding to incoming stimuli. Users are immersed in a competitive game of table tennis against an AI opponent, to which official table tennis rules apply. The difficulty of the AI ranged across five levels from amateur to legendary, ascending in order of difficulty. An increase in difficulty is related to an increase in the speed of serve and return, number of serve/return placements and spin on the ball from the AI. The game utilised haptic, auditory and performance feedback to simulate a real-world environment. For example, when the simulated bat hit the ball users felt a vibration and, realistic sounds were played. Sounds were also used to signal a won or lost point while a scoreboard was also available.
The table tennis setup included a regulation size STIGA table tennis table, Dawei table tennis bats and 40mm Schildkrot table tennis balls. Ten empty standard size soft drink cans were used as serving targets.
In addition to the rallying tasks, serving accuracy was assessed via target accuracy. The targets were ten soft drink cans which were evenly distributed (100mm apart) on the edge of the opposite end of the table. This was used as the target as it is recommended that an ideal serve is one that reaches close to the edge of the opposite end of the table, while bouncing as low as possible off the table. A score was derived based on the number of targets a participant could hit while serving, ranging from 0 (no target hit) to 10 (all targets hit). Each target hit was worth 5 points. A total score was calculated for each participant from 0 to 350 (backhand = 100, forehand = 100, alternating hits = 100, serving = 50). Participants were excluded from the study if they achieved a total score above 90% at pre-test, as it was deemed there was insufficient room for improvement.
As both the quantitative assessment and the quality of skills assessment were assessing different aspects of table tennis skills, it was expected that an individual might improve on one assessment, while not on the other. For this reason, both measures were considered independently in analysis.
There were two participant groups: VR group and control group. All participants underwent a pre- and post-assessment. Participants in the VR group completed a familiarisation (F) session followed by six VR training sessions (T1-T6).
The study hypothesis was addressed using a mixed-model analysis of variance. The analysis comprised a between-subjects (VR training group vs control group) and a within-subjects (pre- and post-training) factor. The dependent variable was real-world table tennis performance assessed by an expert on both quantitative aspects and skill quality aspects.
c80f0f1006