Automated Offside Detection by Spatio-Temporal Analysis of Football Videos

Uchida, Ikuma; Scott, Atom; Shishido, Hidehiko; Kameda, Yoshinari

doi:10.1145/3475722.3482796

Cited by 7 publications

(4 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To avoid the impact of model overfitting, imbalanced data distribution, or using a single background in model training and testing, the first step is to augment the data to expand the dataset. For this purpose, we extracted the video dataset into several randomly chosen images frame by frame from each video during the match [20]. This process resulted in 9,055 images.…”

Section: Results and Analysis 31 Data Preparationmentioning

confidence: 99%

Detecting and Tracking Player in Football Videos Using Two-Stage Mask R-CNN Approach

Husein,

Chalvin,

Ciptady

et al. 2023

conferenceseries

View full text Add to dashboard Cite

Football is one of the most popular sports worldwide and capable of attracting the attention of millions of fans to a single match in the top leagues. The English Premier League, Spanish LaLiga, German Bundesliga, Italian Serie A, and French Ligue 1 are the five best leagues in the world today. There was an experiment where researchers want to analyze the efficiency and accuracy percentage of tracking and detection using the deep learning method of the Mask R-CNN model in classifying positive and negative X-Ray images in football matches. In this study, we applied Mask R-CNN for the segmentation and detection of football players. This model was based on two different backbones, namely ResNet101 and DenseNet. Both backbones produced accuracy values that were not significantly different, but the DenseNet approach performed better than ResNet101 based on testing results in the validation and testing sets. Based on comprehensive experiment results on the dataset, it has been shown that the Mask R-CNN approach with DenseNet can achieve better results compared to Mask R-CNN with ResNet101. Due to insufficient understanding of the characteristics of image types and the uneven distribution of various types of data sourced from random videos, there was still room for improvement in the trained model.

show abstract

Section: Results and Analysis 31 Data Preparationmentioning

confidence: 99%

Detecting and Tracking Player in Football Videos Using Two-Stage Mask R-CNN Approach

Husein,

Chalvin,

Ciptady

et al. 2023

conferenceseries

View full text Add to dashboard Cite

show abstract

“…Foul judgment has also been studied in sports other than scoring competitions. For example, automated offsides detection in soccer [30], automatic detection of faults in race walking [26,27], and detection of dangerous tackles in rugby [18] were proposed. These automated judging systems utilized tracking algorithms and human pose estimation to evaluate athletes' skill as a foul detection based on the competition rules.…”

Section: Related Workmentioning

confidence: 99%

Automatic Edge Error Judgment in Figure Skating Using 3D Pose Estimation from a Monocular Camera and IMUs

Tanaka,

Suzuki,

Takeda

et al. 2023

Proceedings of the 6th International Workshop on Multimedia Content Analysis in Sports

View full text Add to dashboard Cite

Automatic evaluating systems are fundamental issues in sports technologies. In many sports, such as figure skating, automated evaluating methods based on pose estimation have been proposed. However, previous studies have evaluated skaters' skills in 2D analysis. In this paper, we propose an automatic edge error judgment system with a monocular smartphone camera and inertial sensors, which enable us to analyze 3D motions. Edge error is one of the most significant scoring items and is challenging to automatically judge due to its 3D motion. The results show that the model using 3D joint position coordinates estimated from the monocular camera as the input feature had the highest accuracy at 83% for unknown skaters' data. We also analyzed the detailed motion analysis for edge error judgment. These results indicate that the monocular camera can be used to judge edge errors automatically. We will provide the figure skating single Lutz jump dataset, including pre-processed videos and labels at https://github.com/ryota-takedalab/JudgeAI-LutzEdge.

show abstract

“…Other sports use image recognition technology to solve similar problems. Examples include scoring in rhythmic gymnastics (Díaz-Pereira et al, 2014) and figure skating (Xu et al, 2019) and detecting offsides in soccer (Uchida et al, 2021). These studies use competition videos and do not require sensors.…”

Section: Introductionmentioning

confidence: 99%

Automatic Detection of Faults in Simulated Race Walking from a Fixed Smartphone Camera

Suzuki,

Takeda,

Fujii

2024

International Journal of Computer Science in Sport

View full text Add to dashboard Cite

Automatic fault detection is a major challenge in many sports. In race walking, judges visually detect faults according to the rules. Hence, automatic fault detection systems will help a training of race walking without experts’ visual judgement. Some studies have attempted to use sensors and machine learning to automatically detect faults. However, there are problems associated with sensor attachments and equipment such as a high-speed camera, which conflict with the visual judgement of judges, and the interpretability of the fault detection models. In this study, we proposed an automatic fault detection system for non-contact measurement. We used pose estimation and machine learning models trained based on the judgements of multiple qualified judges to realize fair fault judgement. We verified them using smartphone videos of normal race walking and walking with intentional faults in several athletes including the medalist of the Tokyo Olympics. The results show that the proposed system detected faults with an average accuracy of over 90%. We also revealed that the machine learning model detects faults according to the rules. In addition, the intentional faulty walking movement of the medalist was different from that of other walkers. This finding informs realization of a more general fault detection model.

show abstract

Automated Offside Detection by Spatio-Temporal Analysis of Football Videos

Cited by 7 publications

References 14 publications

Detecting and Tracking Player in Football Videos Using Two-Stage Mask R-CNN Approach

Detecting and Tracking Player in Football Videos Using Two-Stage Mask R-CNN Approach

Automatic Edge Error Judgment in Figure Skating Using 3D Pose Estimation from a Monocular Camera and IMUs

Automatic Detection of Faults in Simulated Race Walking from a Fixed Smartphone Camera

Contact Info

Product

Resources

About