Soccer
Soccer field
Bounding Box
Sports
Image
With its wearable EPTS(Electronic Performance Tracking System) technology, Fitogether accurately measures and analyzes the movements of players, and develops B2B solutions for sports teams in making decisions based on big data analytics.
Based on the no.1 wearable technology globally certified by FIFA, Fitogether now provides solutions for about 250 football teams from 35 countries. Including Korea’s K-League, 11 football leagues from 7 countries are EPTS partners, which allows Fitogether to achieve more data.
Recently, the startup has successfully raised Series B investmentworth 10B KRW(8.5M USD). Fitogether aims to expand their technology to video analyzing, and their business to the market of talent identification and fan engagement, using the data collected from performance management.
Datumo provides high quality data for smarter AI. As part of Datumo's Data Sponsorship Program, Datumo cooperated with Fitogether in building the following dataset.
Fitogether currently provides sports IT services using GPS. However, GPS contains its own inaccuracies of not being able to detect layers or depths of the location. Fitogether focuses on developing video solutions for such problems. The purpose of this project is to provide enhanced sports IT solution by locating players in a more accurate manner combining coordinates and GPS information via video data.
The startup aims to develop a video tracker using Datumo’s training data, which is based on various object detection algorithms. Fitogether has also provided diverse data labeling environments in lighting, location, time, uniform, and more. Based on such variety of conditions, the labeling results are anticipated to provide great help in future AI training.
“Taking too much time and having uneven quality were the two toughest issues in labeling data.
Datumo, however, advised us on writing manuals and helped us design roadmaps for making decisions in various possible situations. Datumo's experiences in previous projects helped us to set guidelines for a variety of possibilities which resulted in satisfying quality of data labeling.
We requested quite amount of data and yet received it much earlier than we expected, thanks to Datumo's crowd-sourcing system. The final inspection process was very strict, which was extremely appreciated by our researchers. The finalized data were in desired format to train our AI directly, without additional processing. We would like to thank Datumo with all our hearts for providing much help to bring our project to reality.”
- Image data and JSON files of bbox coordinates and categories of items - 201,835 bbox, 11,150 image data
- Average of 18 bbox per image data
Based on the frame cuts of K-League soccer video clips provided by Fitogether, DATUMO took care of bounding box annotation and labeling of objects from the frame cuts.
- Bounding box annotation and labeling- according to three categories within the soccer video clip
- Datumo's crowd-sourcing platform Cash Mission was used as the main method
- Standards for data preparation were set based on datasets(frame cuts) provided by Fitogether
- Projects, guidelines, and tutorials designed for crowd-workers
- Annotation and inspection of bounding boxes and tags completed by crowd-workers
- Final inspection by in-house workers and delivery of datasets to client
Data were collected and labeled using Cash Mission, Datumo's crowd-sourcing platform.
Sample Data
{
"version": "4.5.6",
"flags": {},
"shapes": [
{
"label": "ball",
"points": [
[
3247.7059959500584,
1058.4013942895517
],
[
3262.863940210376,
1073.5593385498694
]
],
"shape_type": "rectangle",
"labels": [],
"id": 14680567,
"flags": {}
},
{
"label": "others",
"points": [
[
3525.877492464004,
893.5767180596686
],
[
3570.779877095685,
978.3923334750668
]
],
"shape_type": "rectangle",
"labels": [],
"id": 14680568,
"flags": {}
},
{
"label": "players",
"points": [
[
3598.4121137921047,
933.8737299086134
],
[
3632.5686285974007,
1024.8298423676604
]
],
"shape_type": "rectangle",
"labels": [],
"id": 14680569,
"flags": {}
},
{
"label": "players",
"points": [
[
3751.924539883323,
758.4857830993964
],
[
3787.6161789495313,
836.3933393406897
]
],
"shape_type": "rectangle",
"labels": [],
"id": 14680570,
"flags": {}
},
{
"label": "players",
"points": [
[
3309.8666985169993,
757.930618955344
],
[
3344.8792554366382,
832.7147211137965
]
],
"shape_type": "rectangle",
"labels": [],
"id": 14680571,
"flags": {}
},
{
"label": "players",
"points": [
[
3177.974736528456,
859.9089400805065
],
[
3225.5646197201986,
941.1516692435525
]
],
"shape_type": "rectangle",
"labels": [],
"id": 14680572,
"flags": {}
},
{
"label": "players",
"points": [
[
3141.942396397565,
919.7362218072684
],
[
3177.974736528456,
1005.058083815321
]
],
"shape_type": "rectangle",
"labels": [],
"id": 14680573,
"flags": {}
},
{
"label": "players",
"points": [
[
3270.43508101527,
989.7613356465465
],
[
3301.708432826986,
1078.8224027625217
]
],
"shape_type": "rectangle",
"labels": [],
"id": 14680574,
"flags": {}
},
{
"label": "players",
"points": [
[
3040.7330941885934,
1148.8600503071298
],
[
3081.621544619017,
1243.7796673777564
]
],
"shape_type": "rectangle",
"labels": [],
"id": 14680575,
"flags": {}
},
{
"label": "players",
"points": [
[
2924.882484635726,
1019.8667245444833
],
[
2957.4958915266593,
1116.7334106832252
]
],
"shape_type": "rectangle",
"labels": [],
"id": 14680576,
"flags": {}
},
{
"label": "players",
"points": [
[
2847.203980334286,
1642.2220599798945
],
[
2900.273585676678,
1762.8347993944217
]
],
"shape_type": "rectangle",
"labels": [],
"id": 14680577,
"flags": {}
},
{
"label": "players",
"points": [
[
3609.476493434098,
1324.49364357934
],
[
3648.0725700467465,
1433.3897168793133
]
],
"shape_type": "rectangle",
"labels": [],
"id": 14680578,
"flags": {}
},
{
"label": "players",
"points": [
[
3540.5549280543682,
1117.7289474401507
],
[
3576.3941420518277,
1212.1514920103805
]
],
"shape_type": "rectangle",
"labels": [],
"id": 14680579,
"flags": {}
},
{
"label": "players",
"points": [
[
2342.760696101341,
1223.0857057254482
],
[
2379.6681943621866,
1330.9134947620355
]
],
"shape_type": "rectangle",
"labels": [],
"id": 14680580,
"flags": {}
},
{
"label": "players",
"points": [
[
2763.94038213687,
871.378957592687
],
[
2807.3609683261,
928.5493960751729
]
],
"shape_type": "rectangle",
"labels": [],
"id": 14680581,
"flags": {}
},
{
"label": "players",
"points": [
[
2604.007889673207,
751.248669135818
],
[
2635.125976442155,
832.3004300223802
]
],
"shape_type": "rectangle",
"labels": [],
"id": 14680582,
"flags": {}
},
{
"label": "players",
"points": [
[
2203.091143859319,
877.1683690845844
],
[
2247.235406485036,
967.6279236454797
]
],
"shape_type": "rectangle",
"labels": [],
"id": 14680583,
"flags": {}
},
{
"label": "players",
"points": [
[
2072.1057088551424,
935.7861604400446
],
[
2109.736883552475,
1027.6930678739143
]
],
"shape_type": "rectangle",
"labels": [],
"id": 14680584,
"flags": {}
},
{
"label": "others",
"points": [
[
2196.5780559309346,
649.1882063713409
],
[
2229.143495572857,
715.0648473114599
]
],
"shape_type": "rectangle",
"labels": [],
"id": 14680585,
"flags": {}
},
{
"label": "players",
"points": [
[
1401.3748108687116,
988.1678095592739
],
[
1436.1637914889561,
1088.5963762554516
]
],
"shape_type": "rectangle",
"labels": [],
"id": 14680586,
"flags": {}
}
],
"imagePath": "video_01_000037.jpg",
"imageWidth": 3840,
"imageHeight": 2160,
"imageData": null
}
AI regarding locations of football players in the fields
CC BY-SA
Reusers are allowed to distribute, remix, adapt, and build upon the material in any medium or format, even commercially, so long as attribution is given to the creator. If you remix, adapt, or build upon the material, you must license the modified material under identical terms.
https://creativecommons.org/licenses/by-sa/3.0/deed.en