FSD-10: A Dataset for Competitive Sports Content Analysis

Liu, Shenglan; Liu, Xiang; Huang, Gao; Feng, Lin; Hu, Lianyu; Jiang, Dong; Zhang, Aibin; Liu, Yang; Qiao, Hong

doi:10.48550/arxiv.2002.03312

Cited by 4 publications

(7 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…2011-2016 2017-present Football [37]- [42] [30], [43]- [52] Basketball [53]- [59] [60]- [72] Volleyball [73]- [77] [78]-[83] Hockey [84]- [89] [90]-[99] Diving [100] [101]-[107] Tennis [108]- [113] [114]- [123] Table tennis [124]- [129] [130]-[138] Gymnastics [139]- [144] [145]-[148] Badminton [149]- [154] [155]- [164] Figure Skating [165], [166] [2], [167]- [174] Recently, researchers in the communities of computer vision and sports pay much attention to sports video analysis, including building datasets and proposing novel methodologies [2], [17]- [30]. In most existing works on sports video analysis, recognizing the actions of players in videos is crucial.…”

Section: Sportmentioning

confidence: 99%

“…Finally, the 2D TSN that only using RGB frames obtains impressive performance, for example, 61.4% accuracy on FineGym [145] and 87.3% on the generic action recognition dataset -UCF101 [1]. Another variant of TSN is using key video frames instead of random sampling, namely KTSN [2]. Applying key video frames achieves better performance on FSD-10, i.e., 63.3% vs. 59.3%.…”

Section: B Deep Modelsmentioning

confidence: 99%

“…However, in recent sports videos/streaming, the camera motion is no longer fixed and tend to be variant since the highlights of the video keeps changing (e.g., the zoom-in and zoom-out highlights). This naturally leads to the cut of view and more or less occlusion in the recorded videos/streaming, which causes challenges to those well-established action recognition benchmarks [2], [10], [13], [238], [243], [259]- [261] (e.g., those algorithms are barely tolerable to the data sample from different camera motions, with cut and occluded objects). Although there exists work [221], [222], [327] to take the camera motion into considerations when designing the motion descriptor for action recognition task, the cut and occluded objects are still a problem which makes the feature space inconsistent.…”

Section: Challengesmentioning

confidence: 99%

“…Further we showcase two representative datasets, which are table tennis videos (P As we mentioned in Applications section, the action recognition techniques are widely used in web or TV streaming for the purpose of Video Highlights. While the videos are normally recorded via multiple cameras and are in different views [2], [101], [145], [183], [191], [195], [213], this requires the robustness and adaptability of the corresponding action recognition algorithms. Via a thorough investigation in this paper, most of the benchmarks [13], [107], [191], [234]- [239] of action recognition on video datasets focus on single-camera or single-view actions, where it does not conform with the format of sports videos.…”

Section: Challengesmentioning

confidence: 99%

“…Top: common actions in UCF101 [1], which is a coarse annotated dataset for action recognition. Middle: figure skating actions in FSD-10 [2], which is a fine-grained annotated figure skating dataset. Bottom: activities in volleyball, basketball and football [3], where each action could involve multiple players.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

A Survey on Video Action Recognition in Sports: Datasets, Methods and Applications

Wu¹,

Wang²,

Bian³

et al. 2022

Preprint

View full text Add to dashboard Cite

To understand human behaviors, action recognition based on videos is a common approach. Compared with imagebased action recognition, videos provide much more information. Reducing the ambiguity of actions and in the last decade, many works focused on datasets, novel models and learning approaches have improved video action recognition to a higher level. However, there are challenges and unsolved problems, in particular in sports analytics where data collection and labeling are more sophisticated, requiring sport professionals to annotate data. In addition, the actions could be extremely fast and it becomes difficult to recognize them. Moreover, in team sports like football and basketball, one action could involve multiple players, and to correctly recognize them, we need to analyse all players, which is relatively complicated. In this paper, we present a survey on video action recognition for sports analytics. We introduce more than ten types of sports, including team sports, such as football, basketball, volleyball, hockey and individual sports, such as figure skating, gymnastics, table tennis, tennis, diving and badminton. Then we compare numerous existing frameworks for sports analysis to present status quo of video action recognition in both team sports and individual sports. Finally, we discuss the challenges and unsolved problems in this area and to facilitate sports analytics, we develop a toolbox using PaddlePaddle 1 , which supports football, basketball, table tennis and figure skating action recognition.

show abstract

Section: Sportmentioning

confidence: 99%

Section: B Deep Modelsmentioning

confidence: 99%

Section: Challengesmentioning

confidence: 99%

Section: Challengesmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

A Survey on Video Action Recognition in Sports: Datasets, Methods and Applications

Wu¹,

Wang²,

Bian³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

Basketball action recognition based on the combination of YOLO and a deep fuzzy LSTM network

Khobdeh,

Yamaghani,

Sareshkeh

2023

J Supercomput

View full text Add to dashboard Cite

The ability to identify human actions in uncontrolled surroundings is important for human-computer interaction (HCI), especially in the sports areas to offer athletes, coaches, and analysts valuable knowledge about movement techniques and aid referees in making well-informed decisions regarding sports movements. Noteworthy, recognizing human actions in the context of basketball sports remains a difficult task due to issues like intricate backgrounds, obstructed actions, and inconsistent lighting conditions. Accordingly, a method based on the combination of YOLO and deep fuzzy LSTM network is proposed in this paper. YOLO is utilized for detecting players in the frame and the combination of LSTM and fuzzy logic is used to perform the final classification. The reason behind using LSTM along with fuzzy logic refers to its inability in coping with uncertainty which led to the creation of a more transparent, interpretable, and accurate predictive system. The proposed model was validated on SpaceJam and Basketball-51 datasets. Based on the empirical results, the proposed model outperformed all baseline models on both datasets which obviously confirms the priority of our combinational model for basketball action recognition.

show abstract

Full or Weak Annotations? An Adaptive Strategy for Budget-Constrained Annotation Campaigns

Tejero,

Zinkernagel,

Wolf

et al. 2023

2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

View full text Add to dashboard Cite

Annotating new datasets for machine learning tasks is tedious, time-consuming, and costly. For segmentation applications, the burden is particularly high as manual delineations of relevant image content are often extremely expensive or can only be done by experts with domain-specific knowledge. Thanks to developments in transfer learning and training with weak supervision, segmentation models can now also greatly benefit from annotations of different kinds. However, for any new domain application looking to use weak supervision, the dataset builder still needs to define a strategy to distribute full segmentation and other weak annotations. Doing so is challenging, however, as it is a priori unknown how to distribute an annotation budget for a given new dataset. To this end, we propose a novel approach to determine annotation strategies for segmentation datasets, whereby estimating what proportion of segmentation and classification annotations should be collected given a fixed budget. To do so, our method sequentially determines proportions of segmentation and classification annotations to collect for budget-fractions by modeling the expected improvement of the final segmentation model. We show in our experiments that our approach yields annotations that perform very close to the optimal for a number of different annotation budgets and datasets.

show abstract

FSD-10: A Dataset for Competitive Sports Content Analysis

Cited by 4 publications

References 36 publications

A Survey on Video Action Recognition in Sports: Datasets, Methods and Applications

A Survey on Video Action Recognition in Sports: Datasets, Methods and Applications

Basketball action recognition based on the combination of YOLO and a deep fuzzy LSTM network

Full or Weak Annotations? An Adaptive Strategy for Budget-Constrained Annotation Campaigns

Contact Info

Product

Resources

About