Connect with us


Sports video: fine-grained action detection and classification of table tennis strokes from videos




Sports video: fine-grained action detection and classification of table tennis strokes from videos

See the MediaEval 2021 webpage for information on how to register and participate.

Job description

This task offers researchers the opportunity to test their fine-grained classification methods for detecting and recognizing strokes in table tennis videos. (The low variability between classes makes the task more difficult than with common general datasets such as UCF-101.) The task offers two subtasks:

Subtask 1: Stroke Detection: The participants must build a system that detects whether a stroke has been performed, regardless of its class, and infer its temporal limits. The goal is to be able to distinguish between interesting moments in a game (players taking strokes) and irrelevant moments (between strokes, picking up ball, having a break). This subtask can be a preparatory step for later recognizing a performed stroke.

Subtask 2: Stroke Classification: Participants must build a rating system that automatically labels video segments based on a shot taken. There are 20 possible stroke classes.

Compared to Sports Video 2020, this year we expand the task towards detection and also enrich the dataset with new and more diverse stroke samples. The job summary paper is already available here.

Entrants are encouraged to make their code public with their submission.

Motivation and background

Action detection and classification are one of the main challenges in visual content analysis and mining. Sports video analysis has been a very popular research topic due to its variety of application areas, ranging from sports performance analysis and rehabilitation to multimedia intelligent devices with user-tailored summaries. Datasets are now available that focus on sports activities or datasets, including many sports activity classes, and many research contributions are a benchmark for those datasets. A great deal of work is also devoted to fine-grained classification through the analysis of sports gestures using motion capture systems. However, body-worn sensors and markers can disrupt the natural behavior of athletes. In addition, motion capture devices are not always available to potential users, be it a university faculty or a local sports team. It is a challenge to provide end users with the possibility to monitor their physical activities in ecological conditions through simple equipment. The ultimate goal of this research is to produce automatic annotation tools for sports faculties, local clubs and associations to help coaches better assess and advise athletes during training.

Target audience

The task is of interest to researchers in the fields of machine learning, visual content analysis, computer vision, and sports performance. We expressly encourage researchers who specifically focus on domains of computer-aided analysis of sports performance.


Our focus is on footage captured by widespread and inexpensive video cameras, such as GoPro. We use a dataset that is specifically recorded at a sports faculty and is continuously filled in by students and teachers. This dataset consists of player-focused videos recorded in natural conditions without markers or sensors. It consists of 20 table tennis strokes and a rejection lesson. So the problem is a typical video indexing research topic: for a given shot, we need to label the video by recognizing each hit in it. The dataset is subject to a specific usage agreement which can be accessed here.

Evaluation method

Twenty stroke classes are considered according to the rules of table tennis. This taxonomy was designed with professional table tennis teachers. We are working on videos shot in the sports faculty of the University of Bordeaux. Students are the filmed athletes and the teachers guide the exercises performed during the recording sessions. The dataset was recorded in a sports facility with lightweight equipment, such as GoPro cameras. The recordings are unmarked and allow the players to perform in natural conditions from different points of view. These sequences were annotated manually and the annotation sessions were supervised by professional players and educators using a crowdsourced annotation platform.

The training data set shared for each subtask consists of videos of table tennis matches with temporal borders of strokes performed, supplied in an XML file, with the corresponding stroke label.

Subtask 1: Stroke Detection: Participants are asked to temporarily segment regions where a stroke is performed on unfamiliar videos of competitions. The IoU metric on temporal segments will be used for evaluation.

Subtask 2: Stroke Classification: Participants produce an XML file in which each set of test series is labeled according to the given taxonomy. Entries will be judged for accuracy by class and global accuracy.

For each subtask, participants can submit up to five runs. We also encourage participants to perform a failure analysis of their results to understand the mistakes their classifiers are making.

[1] Crisp project

[2] Pierre-Etienne Martin, Jenny Benois-Pineau, Renaud Pteri, Julien Morlier. 2020. Fine-grained recognition of sports actions with conjoined spatio-temporal convolutional neural networks. Multimedia Tools and Applications 79, 2020, 2042920447.

[3] Pierre-Etienne Martin, Jenny Benois-Pineau, Renaud Pteri, Julien Morlier. 3D Attention Mechanism for Fine-Grain Classification of Table Tennis Strokes Using a Twin Spatio-Temporal Convolutional Neural Networks. 2020 25th International Conference on Pattern Recognition (ICPR), 2021, 6019-6026.

[3] Gl Varol, Ivan Laptev and Cordelia Schmid. 2018. Long-term transient convolutions for action recognition. IEEE Trans. Pattern anal. Mach. Intel. 40, 6 (2018), 15101517.

[4] Joao Carreira and Andrew Zisserman. 2017. Quo Vadis, action recognition? A new model and the kinetic dataset. 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017, 4724-4733.

[5] Chunhui Gu, Chen Sun, Sudheendra Vijayanarasimhan, Caroline Pantofaru, David A. Ross, George Toderici, Yeqing Li, Susanna Ricco, Rahul Sukthankar, Cordelia Schmid, and Jitendra Malik. 2017. 2017. AVA: a video dataset of spatio-temporal localized atomic visual actions. 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2018, 6047-6056.

[6] Khurram Soomro, Amir Roshan Zamir and Mubarak Shah. 2012. UCF101: A dataset of 101 classes of human actions from videos in the wild. Computer Vision and Pattern Recognition (cs.CV), CRCV-TR-12-01.

Task Organizers

You can email us directly at

  • Jordan Calandre, MIA, University of La Rochelle, France
  • Pierre-Etienne Martin, Max Planck Institute for Evolutionary Anthropology, Germany
  • Jenny Benois Pineau, Univ. Bordeaux, CNRS, Bordeaux INP, LaBRI, France
  • Renaud Pteri, MIA, University of La Rochelle, France
  • Boris Mansencal, CNRS, Bordeaux INP, LaBRI, France
  • Julien Morlier, IMS, University of Bordeaux, France
  • Laurent Mascarilla, MIA, University of La Rochelle, France

Task Schedule

  • August 1 – October 15, 2021: Data Release
  • October 25, 2021: Expires
  • November 8, 2021: Results back
  • November 22, 2021: Paper work notes
  • December 6-8, 2021: MediaEval 2020 workshop


We would like to thank all players and annotators for their involvement in the acquisition and annotation processes and Alain Coupet from the Sports Faculty of Bordeaux, expert and teacher in table tennis, for the proposed taxonomy of table tennis strokes.




The mention sources can contact us to remove/changing this article

What Are The Main Benefits Of Comparing Car Insurance Quotes Online

LOS ANGELES, CA / ACCESSWIRE / June 24, 2020, / Compare-autoinsurance.Org has launched a new blog post that presents the main benefits of comparing multiple car insurance quotes. For more info and free online quotes, please visit https://compare-autoinsurance.Org/the-advantages-of-comparing-prices-with-car-insurance-quotes-online/ The modern society has numerous technological advantages. One important advantage is the speed at which information is sent and received. With the help of the internet, the shopping habits of many persons have drastically changed. The car insurance industry hasn't remained untouched by these changes. On the internet, drivers can compare insurance prices and find out which sellers have the best offers. View photos The advantages of comparing online car insurance quotes are the following: Online quotes can be obtained from anywhere and at any time. Unlike physical insurance agencies, websites don't have a specific schedule and they are available at any time. Drivers that have busy working schedules, can compare quotes from anywhere and at any time, even at midnight. Multiple choices. Almost all insurance providers, no matter if they are well-known brands or just local insurers, have an online presence. Online quotes will allow policyholders the chance to discover multiple insurance companies and check their prices. Drivers are no longer required to get quotes from just a few known insurance companies. Also, local and regional insurers can provide lower insurance rates for the same services. Accurate insurance estimates. Online quotes can only be accurate if the customers provide accurate and real info about their car models and driving history. Lying about past driving incidents can make the price estimates to be lower, but when dealing with an insurance company lying to them is useless. Usually, insurance companies will do research about a potential customer before granting him coverage. Online quotes can be sorted easily. Although drivers are recommended to not choose a policy just based on its price, drivers can easily sort quotes by insurance price. Using brokerage websites will allow drivers to get quotes from multiple insurers, thus making the comparison faster and easier. For additional info, money-saving tips, and free car insurance quotes, visit https://compare-autoinsurance.Org/ Compare-autoinsurance.Org is an online provider of life, home, health, and auto insurance quotes. This website is unique because it does not simply stick to one kind of insurance provider, but brings the clients the best deals from many different online insurance carriers. In this way, clients have access to offers from multiple carriers all in one place: this website. On this site, customers have access to quotes for insurance plans from various agencies, such as local or nationwide agencies, brand names insurance companies, etc. "Online quotes can easily help drivers obtain better car insurance deals. All they have to do is to complete an online form with accurate and real info, then compare prices", said Russell Rabichev, Marketing Director of Internet Marketing Company. CONTACT: Company Name: Internet Marketing CompanyPerson for contact Name: Gurgu CPhone Number: (818) 359-3898Email: [email protected]: https://compare-autoinsurance.Org/ SOURCE: Compare-autoinsurance.Org View source version on accesswire.Com:https://www.Accesswire.Com/595055/What-Are-The-Main-Benefits-Of-Comparing-Car-Insurance-Quotes-Online View photos


to request, modification Contact us at Here or [email protected]