Dataset for "Exploratory Proof-of-Concept: Predicting the Outcome of Tennis Serves Using Motion Capture and Deep Learning"
Dataset used for training a Machine Learning model to classify tennis serves into “In”, “Out” and “Net” as well as predict the outcome coordinates of the serve. A marker-based motion capture system provided and operated by The University of Bath’s Applied BioMechanics Suite was used to collect spatio-temporal data on participants completing tennis serves, whilst a high-speed video camera recorded the outcome. Dataset has two parts: 1) The spatio-temporal data used for training and validation 2) The serve outcome and coordinates for labelling of the data.
Cite this dataset as:
Durlind, G.,
Martinez-Hernandez, U.,
Assaf, T.,
2025.
Dataset for "Exploratory Proof-of-Concept: Predicting the Outcome of Tennis Serves Using Motion Capture and Deep Learning".
Bath: University of Bath Research Data Archive.
Available from: https://doi.org/10.15125/BATH-01454.
Export
Data
Spatio-Temporal Serve Data.zip
application/zip (21MB)
Creative Commons: Attribution 4.0
Spatio-temporal data collected from an 8-camera marker-based motion capture system consisting of Qualisys Miqus M5 motion capture cameras operating at 180 frames per second with 4 MP (2048×2048) resolution. All marker trajectories were labelled and gap-filled using Qualisys Track Manager. The folder consists of 22 individual serve files which are MAT-files (.mat) and are three-dimensional arrays in the format 75 x 3 x 301:Representing the 75 markers, in a three-dimensional space (X, Y, Z)
Data Labels.xlsx.zip
application/zip (8kB)
Creative Commons: Attribution 4.0
Coordinate and outcome data for each serve, recorded using a high-speed Miqus Video camera spatially aligned and time-synchronised with the motion capture system for simultaneous data collection. The file consists of 22 serves in an Excel spreadsheet, with corresponding names to the spatio-temporal dataset, their outcome (In, Out or Net), and their X and Y coordinates within the service box.
Creators
Gustav Durlind
University of Bath
Uriel Martinez-Hernandez
University of Bath
Tareq Assaf
University of Bath
Contributors
University of Bath
Rights Holder
Documentation
Data collection method:
The methodology can be found in the associated publication.
Additional information:
Further information on the data collection and pre-processing can be found within the associated publication. This includes a detailed experimental set-up, trajectory-filling algorithms for occluded markers, filtering processes for trajectories, and extensive data formatting of spatio-temporal data for compatibility with Machine Learning models.
Documentation Files
Readme.txt
text/plain (585B)
Creative Commons: Attribution 4.0
Readme file
Funders
University of Bath
https://doi.org/10.13039/501100000835
Publication details
Publication date: 14 October 2025
by: University of Bath
Version: 1
DOI: https://doi.org/10.15125/BATH-01454
URL for this record: https://researchdata.bath.ac.uk/1454
Related papers and books
Durlind, G., Martinez-Hernandez, U., and Assaf, T., 2025. Exploratory Proof-of-Concept: Predicting the Outcome of Tennis Serves Using Motion Capture and Deep Learning. Machine Learning and Knowledge Extraction, 7(4), 118. Available from: https://doi.org/10.3390/make7040118.
Contact information
Please contact the Research Data Service in the first instance for all matters concerning this item.
Contact person: Uriel Martinez-Hernandez
Faculty of Engineering & Design
Electronic & Electrical Engineering
Research Centres & Institutes
Centre for Digital, Manufacturing & Design (The Foundry)
