icSPORTS 2023 Abstracts

Area 1 - Computer Systems in Sports
Full Papers Short Papers
Area 2 - Health and Support Technology
Short Papers
Area 3 - Signal Processing in Human Movement
Full Papers
Area 4 - Sport Performance and Support Technology
Full Papers Short Papers

Area 1 - Computer Systems in Sports

Full Papers
Paper Nr: 11
Title:

Automated System for Indoor Sport Video Production
Authors:

Sebastian Purtak, Jagoda Lazarek, Paweł Gora and Łukasz Gąsiorowski
Abstract: Information is the key to success in every area of life, including professional and amateur sports. This is why the recording of matches is so important for sports clubs and organisations, as it allows them to gain an information advantage. In this paper, we present the prototype of the ISVP.AI system, which uses Artificial Intelligence technology to automate the production process of sports recordings in handball and futsal. Our solution involves the use of YOLO family algorithms to detect objects such as players or a ball for each frame of the recording. This information is then used by the expert system, whose task is to detect events relevant to the course of the game. The representation of the state of the game built in this way allows for the automatic production of full sports recordings, match highlights, calculation of statistics and adding effects that improve the attractiveness of the produced recordings. Thanks to this, our system allows for potentially significant reduction of costs associated with recording handball and futsal sports games.
Download
Paper Nr: 12
Title:

Can Ensemble Learning Approaches for Offside Detection Work?
Authors:

Kurt D. Buttigieg, David Suda and Mark A. Caruana
Abstract: The analysis of data collected from various recreational activities and professional sports is essential to obtain more information on the activity in question or to make better data-driven decisions. Most literature related to offside detection related to the efficacy of manual offside detection or the use of an offside detection algorithm. In this study, the focus shall be on the detection of offside judgements in football/soccer using ensemble learning approaches such as random forest type algorithms, boosting type algorithms and majority voting. For random forests, we also consider three corresponding extensions: regularized random forests, guided regularized random forests, and guided random forests. Moreover, five boosting approaches are considered, namely: Discrete AdaBoost, Real AdaBoost, Gentle AdaBoost, Gradient Boosting and Extreme Gradient Boosting. Gentle AdaBoost is the best performing model on most metrics, except for sensitivity, where Extreme Gradient Boosting performs best. Furthermore, soft majority voting among the models considered is capable of improving the Cohen’s Kappa and the F1 score but does not provide improvements on other metrics.
Download
Paper Nr: 16
Title:

The Relationship of Situational Efficiency Parameters of Volleyball Game Phases and Their Intrateam Variability with the Set Score
Authors:

Ivana Klaričić and Zoran Grgantov
Abstract: The purpose of this study is to determine the relationship between situational efficiency parameters of five phases of the volleyball game and their intrateam variability with the set score. A sample of 40 volleyball sets played in the European League for Men in 2011 and 2012 were randomly selected. Although, the sample wasn’t recent, the purpose of this methodologically based study was to propose a new performance indicator. The multiple regression analysis determined a high and positive relationship between the situational efficiency of five phases of the volleyball game with the set score. It also determined that the intrateam variability between the phases of the volleyball game had a statistically significant but negative relationship with the set score. The variability of game phases explained 4.1% of the variance of the score. Conclusion was that a larger negative deviation in situational efficiency of one phase of the game cannot be compensated only by the corresponding increase in another phase of the game, as the linear regression model suggests.
Download
Paper Nr: 18
Title:

VICE: View-Invariant Chess Estimation
Authors:

Kevin Zhu, Alexander Wong and John McPhee
Abstract: A digitized chess match offers chess players a convenient way to study previous matches. However, manually recording a large number of matches can be laborious, while automated methods are usually hardware-based, requiring expensive chessboards. Computer vision provides a more accessible way to track matches from videos. However, current vision-based digitizers are often evaluated on images captured by cameras placed directly above a chessboard, and performance suffers when the camera angle is lower, limiting their applicability. Motivated to develop a more practical solution, we introduce VICE, a view-invariant chess estimator to digitize matches from camera angles not seen during training. Due to its small model size and computational efficiency, VICE is suitable for mobile deployment. By rearranging the framework for chess detection and incorporating prior information from chess and basic geometry, we simplify the chess estimation problem and mitigate the challenges that current chess digitizers struggle with, such as occlusion. We combine the board localization and chess piece detection phases of classical two-step chess estimation to develop a prototype for the first single-step chess digitizer. We show that, with minimal training data, our prototype can infer moves from camera angles that current chess digitizers cannot, while being much smaller in size.
Download
Paper Nr: 23
Title:

Development of a Method for Reproducing Measured Orbital Data of Curling Stone by VR Technology
Authors:

Kouki Ishitoya, Fumito Masui, Hitoshi Yanagi, Michal Ptaszynski and Shimpei Aihara
Abstract: In this study, we report an implementation of reproducing stone trajectory data with VR technology to support curling. The ”Stone Tracking System”, which is used in public curling facilities in Japan, tracks an infrared transmitter attached to a curling stone using an infrared camera to obtain coordinate information and reproduce the trajectory of the stone from the time it is thrown until it stops on a sheet with a two-dimensional representation. The ”Stone Tracking System” was used to measure a set of stone trajectory data. Furthermore, we constructed a process and implemented a database to convert the obtained trajectory shot data from 2D to 3D. Additionally, a curling hole was constructed in a VR space, and a system was built to display the stones, trajectory, and some of the shot information in layers. By using this method, users of the system can project multiple shots on an ice sheet in the VR space, check and compare shot results from any viewpoint, and superimpose shots, which is not possible in real space.
Download
Paper Nr: 24
Title:

Statistical Analysis of Recent Rule Revision Effects for Tactical and Strategic Elements in Curling
Authors:

Yutaka Kondo, Hitoshi Yanagi, Michal Ptaszynski, Shimpei Aihara and Fumito Masui
Abstract: In this paper, PCA is performed on a set of tactical elements calculated by scoring opportunity analysis for the men’s and women’s Japanese Championships 2017, 2022, and 2023 with different rules. Curling is a winter sport called ”chess on ice”. Curling is considered as a system system consisting of multiple elements, which can be divided into three categories: physical, human and tactical. Among these, strategic and tactical factors are considered to be the most important in top-level competition. In addition, in the curling world, some rules are sometimes changed in order to improve players’ skills and tactics. In this study, we analyzed the changes in tactical factors due to the rule changes. As a result, it is considered that the nautical zone rule is not effective in men’s competitions. In the women’s competitions, scoring points by stealing became more important each time the rules were changed.
Download
Short Papers
Paper Nr: 10
Title:

Bayesian Hierarchical Modelling of Basketball Team Performance: An NBA Regular Season Case Study
Authors:

Paul Attard, David Suda and Fiona Sammut
Abstract: The main goal of this study is to propose two Bayesian hierarchical modelling approaches using basketball game data from the 2008/2009 NBA regular season. The aim of the first approach is to estimate the results of each match during the season. This is done by considering each scoring method in basketball separately, that is, free throws, 2-point shots and 3-point shots, and estimating the offensive and defensive ability with respect to each scoring method for each team. These attributes are then used to produce a final score for each match. We attempt both the Poisson and the negative binomial distribution to model the scoring propensities. Both models are used to predict game outcomes and final standings, and since we find the negative binomial approach to be considerably superior, we use it to determine overall attack and defense abilities of each time for each scoring method. The second modelling approach, on the other hand, focuses on finding the probability of the home team winning a particular match in the season. Due to MCMC convergence issues, this model is represented by just one parameter representing overall strength for each team rather than two. When comparing the winning probability approach with the scoring propensity approach, we find that the latter is superior at predicting game outcomes, the former is superior at predicting final standings, while both are comparable in predicting which teams will qualify to playoffs.
Download
Paper Nr: 13
Title:

Detection of Shot Information Using Footwork Trajectory and Skeletal Information of Badminton Players
Authors:

Naoki Tanaka, Hidehiko Shishido, Masashi Suita, Takeshi Nishijima, Yoshinari Kameda and Itaru Kitahara
Abstract: As video analysis has become important for sports science, various research has been conducted. In badminton, while shot information is essential primary data for performance analysis, it has been input manually, which makes it difficult to give instant feedback onsite. Our research aims to automatically detect shot information from videos of badminton game. By applying video tracking, the player’s footwork trajectory and skeletal information are estimated. Based on the estimated information, the hit timing is detected using deep learning classification. The horizontal position of the hit point, which is useful for game analysis, is also detected from the player’s footwork trajectory around the hit timing.
Download
Paper Nr: 14
Title:

Vertical Launch Angle Measurement of a Golf Ball Using Audio and Monocular Video Data
Authors:

Jan-Bernd Menke-Zumbrägel, Anna Brinkmann and Andreas Hein
Abstract: n golf, several parameters can be measured that describe how the golf ball was hit and how the ball lifts after impact with the golf club, the so-called launch parameters. In addition to the spin rate or the velocity of the ball and club head, the launch angle is an important value that describes the vertical component of the ball’s launch direction. For professional use, there are systems called launch monitors that use either radar-technology, multiple high-speed cameras or a combination of both to measure the above parameters. Despite their high accuracy, these systems can suffer from disadvantages such as some inconvenience regarding size, weight or setup, and an inaccessibility regarding high cost. Therefore we present a method for vertical launch angle measurement based on monocular low frame rate video and audio data, by detecting the motion blur structure created by the launching golf ball. This approach allows the vertical launch angle to be measured with a simple and inexpensive setup that achieves an accuracy of ± 0.74°, which is comparable to a commercial launch monitor.
Download
Paper Nr: 17
Title:

2Trax3: Raising Accessibility and Everyday Use of Automatic Motion Analysis in (Combat) Sports via ML Enhanced 2D to 3D Estimation Algorithms
Authors:

Samir Duvelek, Dominik Hoelbling, René Baranyi, Roland Breiteneder, Karl Pinter and Thomas Grechenig
Abstract: A sound technique forms the fundamental basis for many sports, particularly Martial Arts, as it often distinguishes between successful hits and being hit. However, the process of improving one’s technique is highly intricate, often requiring expert feedback and expensive technology such as 3D motion capturing. The integration of automated technique analysis has the potential to streamline this process and make it more accessible. In this study, the aim is to democratize technique analysis by developing and evaluating a web application. This application allows users to upload 2D video recordings of themselves performing the double side kick technique and receive immediate feedback. To validate the analysis generated by the application, it was compared to a Vicon motion app 3D analysis of the same data from a preliminary study involving 44 participants. The results of Bland-Altman plot analysis demonstrated a highly significant agreement between the 3D and 2D performance indicators (Mean differences: relative phase duration: <0.04s; vector spreading angle: <15 degrees; relative body position <13%), indicating that the web application is a suitable tool for fast and effective motion analysis.
Download
Paper Nr: 25
Title:

Development of a Curling Stone Tracking System Using Infrared LEDs, and an Accompanying Application
Authors:

Yoshinari Takegawa, Noa Sasaki, Shimpei Aihara and Fumito Masui
Abstract: The purpose of this study is the design and implementation of a real-time position measurement system for use on curling stones. Often called ‘chess on ice’, curling is a sport that requires a high level of strategy. Accordingly, how the stones move around the vast 40m curling rink constitutes important data. However, in the unique environment of the icy and vast rink, it is difficult to monitor the position of the stones by a simple method without hindering the players. Therefore, in our research, we proposed a system using infrared LEDs and an infrared camera. Infrared LED modules are installed on the stones and the rink, and infrared cameras installed around the edge of the rink film the LED modules and perform calibration. Then, using four coordinates of the LED modules on the rink, the system employs perspective transformation technology, which is a type of image processing. In so doing, it is possible both to measure the position of the stones, and solve problems. Through experiments, performance evaluation was conducted to asses what degree of error occurs in position measurement when the proposed system is used. Experiments were conducted on a curling rink. The average error was 0.189m in the experiment at the curling rink.
Download
Paper Nr: 36
Title:

Development of Monocular Vision-Based Tracking Method for Wheelchair Sports
Authors:

Shimpei Aihara, Takara Sakai and Akira Shionoya
Abstract: Recently, tracking systems to measure player positions have been introduced in the sports domain. However, wheelchair sports have not been considered extensively. In addition, user-friendly and low-cost systems for wheelchair sports are uncommon. Thus, in this paper, we propose a method to calculate the kinematic data of wheelchair athletes on a playing field (i.e., player positions and wheelchair directions) using images acquired by a monocular camera. The proposed method was evaluated experimentally, and the root mean square error of the position accuracy was 0.11 m, and the mean average error of the direction accuracy was 6.78 degrees. The results demonstrate that the proposed method outperforms existing tracking methods in terms of accuracy. The findings of this study suggest that it is possible to acquire kinematic data of wheelchair athletes using a simple method, which we expect to contribute to improvement analysis of the wheelchair athlete performance.
Download
Paper Nr: 44
Title:

FOOTBALLTrace: An AI-Based System for Football Player Tracking with Occlusion Detection and Trajectory Correction
Authors:

Abdelrahman H. Mostafa, Muhammad A. Rushdi, Tamer A. Basha and Khaled Sayed
Abstract: Data analytics have had a significant impact on tactical and workload planning in football. Football data is divided into two categories: event data, which captures on-the-ball events like passes and shots, and tracking data, which captures off-the-ball movements. However, traditional methods of collecting tracking data are expensive and inconvenient. Recently, AI solutions have emerged as low-cost and user-friendly alternatives to track players’ movements from video streams. This paper introduces FOOTBALLTrace, an end-to-end AI system for tracking multiple football players from a panoramic game view. The system also incorporates a novel algorithm that detects potential occlusion events and ensures trajectory continuity for occluded players. The workflow involves five stages: panoramic view creation, player detection, player ID association, occlusion detection, and trajectory correction. The system utilizes YOLOv7 for multiple object detection and employs a pre-trained deep affinity network to assign unique IDs to players throughout the game. Occlusion detection and trajectory correction are achieved by extracting geometric features from discontinuous player trajectories. The system’s performance was evaluated on full-length video data of a football game, with occlusion events manually extracted for training and testing the occlusion detection and trajectory correction algorithm. The system achieved an 87.5% trajectory correction rate for occluded trajectories.
Download
Paper Nr: 8
Title:

Validation of a Biomechanical Performance Assessment Platform Applying an Inertial-Bpased Biosensor and Axis Vector Computation
Authors:

Wangdo Kim, Sean S. Kohles, Emir A. Vela and Victor Huayamave
Abstract: Inertial kinetics and kinematics have substantial influences on human biomechanical function. A new algorithm for IMU-based motion tracking is presented in this work. This study combines recent developments in improved biosensor technology with mainstream motion-tracking hardware to measure the overall performance of human movement based on joint axis-angle representations of limb rotation. This study proposes an alternative approach to representing three-dimensional rotations using a normalized vector around which an identified joint angle defines the overall rotation, rather than a traditional Euler angle approach. Contrast the procedure of Euler angles with the procedure of Axis angle, Euler angles force the body to move along a certain route which it had arbitrarily chosen but which the body had not chosen; in fact, the body would not take any of its routes separately, though it would take all of them together in the most embarrassing manner-goal-directed behavior. But axis angle had no preconceived scheme as to the nature of the movements to be expressed. Although the axis-angle representation requires vector quotient algebra (quaternions) to define rotation, this approach may be preferred for many graphics, vision, and virtual reality software applications. Elbow flexion and extension motion was used to validate the analytical methods. The results suggest that the novel approach could reasonably predict a detailed analysis of axis-angle migration. The described algorithm could play a notable role in the biomechanical analysis of human joints and offers a harbinger of IMU-based biosensors which may assess the control of skilled manipulation.
Download
Paper Nr: 31
Title:

Construction of a Virtual Environment to Measure the Evolution of Kendo Athletes
Authors:

Francisco M. A. de Araújo, Antonio C. Ferreira, Matheus A. Dantas, Halyson C. Pimentel, Paulo A. Leal, Sérgio L. B. de Carvalho, N. F. Ferreira, António Valente and Salviano P. Soares
Abstract: The use of technology applied in sports comes each year becoming a great tool to help athletes train. Moreover, the post-pandemic world is undergoing dramatic changes in the way of thinking and acting, with new ways of exercising emerging, but without leaving home. Thus this paper describes the development of a platform for training, focusing on Kendo practitioners (Japanese fencing) using virtual reality tools to allow athletes and training the distance. Through the use of a HMD (Head Mounted Device), kendōkas will be able to practice blows and improve their reflex by a gamified experience in a virtual environment.
Download
Paper Nr: 32
Title:

Evaluating the Impact of Mouse Curvature Design on Hand Ergonomics and Comfort in E-Sport Players Through EMG and 3D Kinematics Measurement Methods
Authors:

Wen-Tzu Tang
Abstract: This research compares the impact of two different mouse curvature designs on comfort parameters, wrist muscle fatigue, and E-sport performance. Sixteen elite male right-handed E-sport players participated in the study, testing five different mouse models. Click performance, muscle activation signals, finger pressure, and 3D motion data were recorded. The testing involved five different mouse models (BenQ Zowie, FK1, FK1+, FK2, S1, S2) to compare hand ergonomics and comfort, with peak position and length as the parameters. Click performance was tested. Analysis revealed no significant difference in click performance between mouse models, but FK1 showed reduced fatigue after 60 seconds of clicking. Different curvatures resulted in varying wrist angles and fatigue levels. The results showed that bigger length and height on the back for symmetrical mouse did not provide a performance advantage, with the back tall design exhibiting higher fatigue. The mid-tall design showed less fatigue and could potentially reduce wrist extension and stress on the carpal tunnel. This study provides insights into the impact of mouse curvature on hand ergonomics, comfort, and muscle fatigue in E-sport players, aiding mouse design improvements and consumer guidance.
Download

Area 2 - Health and Support Technology

Short Papers
Paper Nr: 26
Title:

Accelerometer Based Body Movement Quantification in Classroom Lectures: Seated Activity Comparison Between Body Regions
Authors:

Muhammad U. Khan, Francesca Gallè, Giada Ballarin, Patrizia Calella, Giuseppe Cerullo, Giorgio Liguori and Giuliana Valerio
Abstract: Sitting behavior research rarely consider non-ambulatory movement in separate body regions. This study used accelerometers, a sedentary cut off criterion, and measurement variables to evaluate movement accumulation in trunk, waist, and foot regions of students in a 42-minute classroom session. Findings show that all three sites were unique in stationary and movement measures (P≤0.012). Trunk and waist spent almost entire lesson pe-riod in stationary state (>98%) whereas foot spent larger proportion in movement (9%). In addition, longest stationary period in trunk and waist regions exceeded the 30-minute threshold of prolonged sitting by a margin of 1 to 2 minutes as opposed to the foot. Altogether, trunk and waist recorded negligible seated activity and foot recorded sporadic and frequent movement. Based on health connection of body regions movement while sitting, we believe that some movement may be better than no movement at all. Since trunk and wait were inac-tive during the lesson period, strategies could be established to encourage intermittent movement in static body regions and facilitation of movement in already active regions. However, further investigation is needed to bet-ter understand dependencies of localized body activity on students’ wellbeing in prolonged sessions of class-room lessons.
Download
Paper Nr: 49
Title:

No Pain no Game or More Game More Pain? The Effects of Summer Congested Games Periods on Professional European Male Football Players’ Injury Occurrences for the Following Season: A Data Analytics Approach
Authors:

Andria Procopiou and Koulla Parpa
Abstract: Two recent congested periods, the post Covid19 lockdown period (May-August 2020) and Euro 2020 (JuneJuly 2021) forced European male footballers to consistently play games for nearly three years with long duration of congested games and insufficient rest days. This phenomenon has not been previously observed to such extent. This study uses data analytics and statistics to investigate how these two congested periods during seasons 19/20, 20/21 and 21/22 affected the injuries occurrences in European male footballers who both participated in Euro 2020 and played in one of the seven major European leagues, using Transfermarkt data. The results indicate that the severity and injury days out for each injury were significantly increased when comparing the 20/21 and 21/22 seasons (before and after Euro 2020), but not for 19/20 and 20/21 seasons. In contrast, there were no significant differences on injury rates and occurrences when comparing the congested games period after March 2020 lockdown with the season before lockdown or the 20/21 season. Furthermore, the players whose national teams reached the group of 16 stage onwards in Euro 2020, exhibited more fitness-related occurrences during the 21/22 season. With regards to the injuries each player exhibited, there was a significant difference on their total days out, and days out for their first and longest injury when comparing seasons 20/21 and 21/22. Also, there was a significant increase on the severity of injuries each player exhibited as well as the how soon each player was reported with an injury. However, no similar results were observed when comparing the 19/20 season (before the Covid19 March 2020 lockdown) and the 20/21 season.
Download
Paper Nr: 53
Title:

Influence of Sport on Autonomic Dysreflexia of a Patient with Spinal Cord Injury
Authors:

Laura Bellintani de Freitas, Rafaella Camilo de Oliveira, Bruna V. Zuchatti, Ed W. Ferrari Junior, Orcizo F. Silvestre and Alberto Cliquet Junior
Abstract: Spinal Cord Injury (SCI) causes loss of motor, sensory and autonomic functions below the injured level. The increase in the number of cases of SCI, the main cause being motor vehicle accidents, and the social impact that this condition causes makes the study on this condition very relevant. SCI brings physiological changes and physical adaptations to the individual, such as cardiovascular problems and increased blood pressure due to lacking of the sympathetic nervous system, known as autonomic dysreflexia (AD). Topographic observation of the retina can help to understand the change that occurs in the autonomic response of these individuals. The objective of the study was to analyze the vasculature changes in SCI with AD in athletes and non-athletes, through retinal photography of these individuals. Retinal photographs were taken of 40 subjects participating in the research, and these were divided into 4 groups: non-athletes with AD, athletes with AD, SCI without AD, individuals without SCI. We can conclude that there is a higher prevalence of vascular changes in patients with AD, especially in the athlete group.
Download
Paper Nr: 28
Title:

Analysis of the Reaction Time and Dominance in Elderly Men: A Pilot Study
Authors:

Francesca Campoli, Vincenzo Bonaiuto, Lucio Caprioli, Saeid Edriss, Emilio Panichi, Michele Panzarino, Cristian Romagnoli, Giuseppe Annino and Elvira Padua
Abstract: The study analyses reaction times in elderly subjects and investigates the reactivity of the dominant and non-dominant hand and foot. 20 men (73.3 ± 3.1 years), healthy, free from injury and without any physical problem that could affect the test results. The tests are the baseline reaction times, the plate tapping test and foot tapping test. Descriptive statistical procedures are presented as mean ± SD and the percentage changes (Δ%) were calculated for each test. The significance level was P<0.05. In baseline reaction time test, the non-dominant hand showed a shorter reaction time than the dominant hand of about 28 msec (p = 0.05). In the plate tapping test, on the other hand, the dominant hand was more reactive than the non-dominant hand (7%, p = 0.002). In the foot test, the difference between the dominant and non-dominant feet was 3% (p = 0.1). The difference in performance between the tests performed could be due to the difficulty of the required motor task. The study lays the foundation for developing a motor work protocol focused on reactivity, a motor ability that physiologically degenerates with age and is of fundamental importance for the individual's physical and cognitive well-being.
Download
Paper Nr: 34
Title:

Energetic Cost of Running in Track and Treadmill
Authors:

Carlo M. Biancardi, Leonardo Lagos-Hausheer, Germán Pequera, Enzo Castroman, Federico Cazot, Enzo Martinez and Renata L. Bona
Abstract: The metabolic power and cost of running per unit distance on a track have been estimated and compared with data collected indoor, in a laboratory on a treadmill. Oxygen uptake have been collected using a portable device, while speed was regulated by auditory feedback (metronome) and verified using GPS. Speed fluctuations remained within an acceptable range. Metabolic power increased linearly with speed, with a slope significantly lower on the track than on the treadmill (p = 0.017). However, statistical comparisons at the same speed did not yield significant differences between the two conditions. The average cost of transport was slightly, but not significantly, lower on the track (4.20 J/kg/m) than on the treadmill (4.35 J/kg/m), and it remained nearly independent of speed over a wide range. Nevertheless, in the lower and higher speed ranges on the track, the cost of transport tended to increase. A similar non-linear trend was observed in the cost of transport in relation to step frequency, with the minimum values falling within a range of 160 to 180 steps per minute. These preliminary results are encouraging and warrant further research to explore the differences between running on a treadmill and on a track.
Download

Area 3 - Signal Processing in Human Movement

Full Papers
Paper Nr: 7
Title:

Reaction Time Estimation Based on Recursive Short-Term Principal Component Analysis for Skeletal Information of Badminton Players
Authors:

Kana Sagawa, Hidehiko Shishido, Masashi Suita and Itaru Kitahara
Abstract: The aim of this paper is to measure the shot-reaction intervals of badminton players based on time-series 3D skeletal information. In competitions where game dominance changes, effective plays and tactics in situations can be investigated by analyzing the measured reaction intervals. In our proposed method, we estimated shot-reaction intervals using a badminton player’s motion information and applied a short-term principal component analysis to the sequential 3D skeletal information of athletes to extract features useful for motion analysis. Hit and reaction times were detected by identifying the extrema in the first and second principal component scores. We estimated a shot’s reaction interval from the hit time to the reaction time at which the player starts moving in response. We applied the proposed method to the 3D skeletal information of a badminton player and confirmed that reaction intervals can be estimated. By using the results of this study to provide feedback to badminton players on the analysis of reaction intervals, players can learn and improve their effective and ineffective tactics and plays.
Download

Area 4 - Sport Performance and Support Technology

Full Papers
Paper Nr: 29
Title:

Real-Time Karting Performance Monitoring via DAQ System with RTK-Enhanced GPS
Authors:

Saeid Edriss, Paolo Boatto, Giuseppe Annino, Francesca Campoli, Lucio Caprioli, Nunzio Lanotte, Emilio Panichi, Cristian Romagnoli and Vincenzo Bonaiuto
Abstract: Measuring instantaneous speed and engine parameters and accurately assessing trajectories is paramount in evaluating racing performance in motorsports. DAQuino is a general-purpose acquisition system that can configure these specific sports applications. This paper deals with an application of such a DAQ, based on RTK-enhanced GPS, suited for the telemetry acquisition of a kart’s parameters and driver performance assessment. The proposed system measuring the kart’s position with a maximum error of a few centimetres can assess the effectiveness of the trajectories.
Download
Short Papers
Paper Nr: 27
Title:

Video Analysis Application to Assess the Reaction Time in an ATP Tennis Tournament
Authors:

Lucio Caprioli, Francesca Campoli, Saeid Edriss, Elvira Padua, Emilio Panichi, Cristian Romagnoli, Giuseppe Annino and Vincenzo Bonaiuto
Abstract: 2D Video analysis is often used in tennis to analyze the players' technique or issues related to game tactics. This paper applies video analysis to assess the reaction time in tennis matches. Fifteen subjects were examined (26.20 ± 4.75 years old, weight 79.13 ± 5.67 kg, height 184.40 ± 5.30 cm, BMI 23.26 ± 1.19), all with an ATP ranking between the #130 position and the #1066 position updated on the day of the sampling. The average RT was 0.248 ± 0.07 s. The longer reaction times were recorded at the first stroke after the serve, while the shorter were in defensive situations when the opponent was attacking or playing a volley, and the examined player often anticipate by starting the mouvement even before the opponent's stroke. The reaction times of high-level tennis players were found to be very short, often less than 120ms in defensive actions. These results prompt us to consider the importance of kinetic perceptual skills such as reaction speed and anticipation in tennis training.
Download
Paper Nr: 43
Title:

Modelling Drag Forces on a Wheelchair Racing Simulator
Authors:

Ateayeh Bayat and Félix Chénier
Abstract: A racing wheelchair simulator is a stationary device that allows para-athletes to train inside, in different simulated conditions (e.g., rolling resistance, wind). Apart from improving performance during training sessions, it also allows researchers to study the biomechanics of the propulsion in a laboratory, which is important because wheelchair racing can cause musculoskeletal disorders and pain in athletes. For these reasons, the realism of these simulators is paramount, and having a model that reproduces a non-linear relationship between drag force, wheeling speed and wind speed is one of the criteria for achieving this realism. In this work, we develop and characterize such a model through empirical data recorded on a racing track, and then implement it on a racing wheelchair simulator with a haptic controller. Propelling on this simulator proved to be somewhat less challenging than real-life conditions, with a measured resistance force approximately 5 N lower than during real propulsion conditions. However, the resistance expectedly increased as simulated face wind increased. These results show a promising avenue for racing wheelchair athletes, both for training and assessing/correcting their biomechanics.
Download
Paper Nr: 50
Title:

The 12th Player: Explainable Artificial Intelligence (XAI) in Football: Conceptualisation, Applications, Challenges and Future Directions
Authors:

Andria Procopiou and Andriani Piki
Abstract: Artificial intelligence (AI) has demonstrated tremendous progress in many domains, especially with the vast deployment of machine and deep learning. Recently, AI has been introduced to the sports domain including the football (soccer) industry with applications in injury prediction and tactical analysis. However, the fact remains that the more complex an AI model is, the less explainable it becomes. Its black-box nature makes it difficult for human operators to understand its results, interpret its decisions and ultimately trust the model itself. This problem is magnified when the decisions and results suggested by an AI model affect the functioning of complex and multi-layered systems and entities, with a football club being such an example. Explainable artificial intelligence (XAI) has emerged for making an AI model more explainable, understandable and interpretable, thus assisting the creation of human-centered AI models. This paper discusses how XAI could be applied in the football domain to benefit both the players and the club.
Download
Paper Nr: 55
Title:

Investigations of the Throwing Biomechanics Index in Collegiate Baseball Pitchers
Authors:

Hannah Stokes, Koco Eaton and Nigel Zheng
Abstract: In the field of sports biomechanics the aim is to improve performance and reduce injury. In this study we create a novel throwing biomechanics index by using logistic regression to identify the most important and significant variables that influence injury. Fifteen biomechanics (kinematic and kinetic) variables were identified using logistic regression and the standards of the throwing biomechanics index were determined based on the healthy and high performing group (ball speed > 80 mph). Z-scores were used to determine the index value for each pitcher. Division 1 and 2 collegiate baseball pitchers participated in this study that were grouped based on their injuries before and after the study compared to the healthy group. The healthy group had the highest throwing biomechanics index and further analysis will provide more insights on both injury and performance. The throwing biomechanics index found significant relationships with the pitcher’s height (p=.0165), mass (p=.0003), age (p=.0099), forearm length (p=.0001), internal flexibility (p=.0015), external ROM (p=.0002), and external flexibility (p=.0142). There is great value in quantifying a throwing biomechanics index for both understanding the injury mechanisms and for improved performance.
Download