This study presents a deep learning model devoted to the analysis of swimming using a single Inertial Measurement Unit (IMU) attached to the sacrum. Gyroscope and accelerometer data were collected from 35 swimmers with various expertise levels during a protocol including the four swimming techniques. The proposed methodology took high inter- and intra-swimmer variability into account and was set up for the purpose of predicting eight swimming classes (the four swimming techniques, rest, wallpush, underwater, and turns) at four swimming velocities ranging from low to maximal. The overall F1-score of classification reached 0.96 with a temporal precision of 0.02 s. Lap times were directly computed from the classifier thanks to a high temporal precision and validated against a video gold standard. The mean absolute percentage error (MAPE) for this model against the video was 1.15%, 1%, and 4.07%, respectively, for starting lap times, middle lap times, and ending lap times. This model is a first step toward a powerful training assistant able to analyze swimmers with various levels of expertise in the context of in situ training monitoring.

This paper presents a novel sensor-to-segment calibration procedure for inertial sensor-based knee joint kinematics analysis during cycling. This procedure was designed to be feasible in-field, autonomously, and without any external operator or device. It combines a static standing up posture and a pedaling task. The main goal of this study was to assess the accuracy of the new sensor-to-segment calibration method (denoted as the ‘cycling’ method) by calculating errors in terms of body-segment orientations and 3D knee joint angles using inertial measurement unit (IMU)-based and optoelectronic-based motion capture. To do so, 14 participants were evaluated during pedaling motion at a workload of 100 W, which enabled comparisons of the cycling method with conventional calibration methods commonly employed in gait analysis. The accuracy of the cycling method was comparable to that of other methods concerning the knee flexion/extension angle, and did not exceed 3.8°. However, the cycling method presented the smallest errors for knee internal/external rotation (6.65 ± 1.94°) and abduction/adduction (5.92 ± 2.85°). This study demonstrated that a calibration method based on the completion of a pedaling task combined with a standing posture significantly improved the accuracy of 3D knee joint angle measurement when applied to cycling analysis.

Match outcomes and championship rankings are the consequence of the team’s technical, tactical, and physical variables. This study aimed to compare physical and modern technical-tactical performance variables between matches with different outcomes for a professional soccer team. Total distance covered, distance covered between 20.0 and 25.0 km/h, and at > 25.0 km/h, distance covered at ≥ 3 m/s2 and at ≤ −3 m·s−2 along with several modern technical-tactical variables (expected goals in favor (xG) and against (xGA), expected goals chain in favor (xGC) and against (xGCA) and passes per defensive action (PPDA)) were collected for 71 soccer matches during the 2020/2021 and 2021/2022 seasons from a team competing in the French Ligue 2. These technical-tactical and running performance variables were obtained by a validated video tracking system (OPTA) and their values per match were compared depending on the match outcome which was categorized as “loss” (L), “draw” (D) or “win” (W). No significant differences were observed for the different running metrics depending on the match outcome. However, significant differences were observed for xGA (0.70  0.39 vs. 1.24  0.59; pbonferroni= 0.004) and xGCA (5.38  2.78 vs. 10.92  6.18; pbonferroni = 0.002) between wins and losses, respectively. Additionally, there was a weak, but significant correlation between xGCA and distance covered in acceleration (r = 0.255; p = 0.032) and deceleration (r = 0.237; p = 0.047). In conclusion, while our study found associations between technical-tactical variables and match outcomes, causality cannot be inferred. Improved technical-tactical performance may positively impact the match result, especially by the reduction of the opposing team's goal expectancy. On the contrary, running performance variables showed no associations with the match outcome.

Prospective decision-making in soccer requires taking information about teammates and opponents while ensuring effective reception and control of the ball. This multitasking aspect of soccer has rarely been considered in laboratory studies examining the relationship between expertise and visual attention measured in a visual tracking task. This research explored the ability of soccer players to pay attention to their environment when receiving a ball in simulated backward passing situations in a CAVE. Twenty-four soccer players, 12 high-skilled and 12 less-skilled, participated in this study. The visual tracking performance and global displacement of both groups were compared across different experimental conditions, which involved monitoring virtual teammates and opponents, receiving a virtual ball, or performing both tasks simultaneously. When the virtual ball was present in the scene, visual tracking performance decreased, and the global displacement increased. However, no additional differences were observed when ball reception was required. Furthermore, no differences between the two groups of participants were observed under any experimental condition. Adding the possibility to interact with the ball appears to engage participants in movement, introducing a constraint that exists in real game situations when perceiving teammates and opponents. However, the relationship between the perceptual-cognitive skills involved in visual tracking and expertise in soccer remains unclear. Further adaptations to the experimental protocol and the virtual environment are required to thoroughly investigate soccer players’ ability to interact with their complex and dynamic performance environment.