SlideShare ist ein Scribd-Unternehmen logo
1 von 24
9th International Conference on Computer and Knowledge Engineering (ICCKE 2019), October 24-25 2019, Ferdowsi University of Mashhad
Tennis stroke detection
using inertial data of a smartwatch
Sara Taghavi, Fardjad Davari, Hadi Tabatabaee Malazi, Ahmad Ali Abin
Faculty of Computer Science and Engineering, G.C.
Shahid Beheshti University, Tehran, Iran
Motivation
• Analyzing sports activities are changing:
• Question is:
How accurate a commercial smartwatch can detect tennis strokes?
122
Outline:
• Introduction
• Related works
• Proposed method
• Evaluation scenarios:
i. Effect of applying proposed method on two datasets
ii. Proposed method classification results for smartwatch dataset
iii. Classification results & classification accuracy improvement
Introduction
Related works
Proposed Method
Evaluation scenarios
222
Introduction
• Smartwatch for Tennis Activity Detection -> its popularity
• As it is not an open eco-system ->
Challenges:
1. Fluctuating sampling rate of sensor data
2. Limited data transmission rate
3. Highly resourced constraint regarding memory space
4. The processing power adds more constraints
Introduction
Related works
Proposed Method
Evaluation scenarios
322
Network point of view
Introduction
Related works
Proposed Method
Evaluation scenarios
422
Related works
• M. Kos et al. introduced a system:
designed a lightweight device
can be worn on the wrist
communicates with the personal computer (PC) through USB
• Pei et al. in developed a product:
Has three modules : sensor, controller and transmission
is embedded in tennis racket
the motion information is sent to the mobile phone
• Dhnesh et al. introduced a platform:
Wireless measurement sensors
attached to the racket and player’s body
work in conjunction with software analysis modules on PC
Introduction
Related works
Proposed Method
Evaluation scenarios
522
Related works
• Previous works in Tennis Activity Detection -> custom-built wearables
Advantages:
-> Programmable micro-controllers
-> Capable of adjusting sufficient memory space
An open customizable eco-system
High-quality datasets
Very few of them are publically available!
Introduction
Related works
Proposed Method
Evaluation scenarios
622
Proposed Method
• Phase one
-> The Data Collector Application on the smartwatch
• Phase two
-> Data pre-processing and Tennis Activity Detection
DATA TAD
Pre-processed
Data
ML methods
Introduction
Related works
Proposed Method
Evaluation scenarios
722
Phase I: Data Collector Application
1. Haptic feedback
2. Logger
3. Sensors helper
4. Messaging helper
5. Client SDK
Introduction
Related works
Proposed Method
Evaluation scenarios
822
Smartwatch application is designed with 5 modules:
Phase I: Data Collector Application
The most important modules are:
1. Sensor helper
2. Messaging helper
3. Client SDK
922
Introduction
Related works
Proposed Method
Evaluation scenarios
Data Collection Procedure
1022
Introduction
Related works
Proposed Method
Evaluation scenarios
Data sets characteristics
• Smartwatch data set:
-> Sensors: 3D Accelerometer & 3D Gyroscope(angular velocity)
-> Sampling rate: 50Hz
-> 8 subjects, 4 try for each stroke
-> Strokes : Serve, Forehand, Backhand
Same as:
• UTD-MHAD multi-modal data set:
-> Strokes: Serve , Forehand
-> Custom-built wearable inertial unit
-> Publicly available
Related works
Introduction
Proposed Method
Evaluation scenarios
1122
Introduction
Proposed Method
Evaluation scenarios
Phase II: Data pre-processing & Activity Recognition
-> To elevate data quality
-> Reduce effects of poor data quality for AR process
Steps Include:
• Data Cleansing -> exclude irrelevant data
• Stream Alignment -> equal number of windows containing data points
• Signal Processing -> to reduce noise -> high-pass and low-pass filters
Related works
Introduction
Proposed Method
Evaluation scenarios
1222
Introduction
Proposed Method
Evaluation scenarios
Phase II: Smartwatch Signal Processing
Related works
Introduction
Proposed Method
Evaluation scenarios
1322
Related works
Introduction
Proposed Method
Evaluation scenarios
Phase II: UTD-MHAD Signal Processing
Related works
Introduction
Proposed Method
Evaluation scenarios
1422
Related works
Introduction
Proposed Method
Evaluation scenarios
Phase II: Feature extraction & Feature importance
• An 18 dimension data matrix for each data steam
• Segmentation method: Fix time-based sliding window with overlap
Related works
Introduction
Proposed Method
Evaluation scenarios
1522
Related works
Introduction
Proposed Method
Evaluation scenarios
Features: Feature importance:
Tools
• Smart watch device :
Fitbit Ionic™ Watch
• Phase I: Data Collector Application development
-> IDE: Fitbit online Studio
-> JavaScript
-> compile, bundle and optimize -> TypeScript compiler and rollup.js
-> JavaScript is run on the device using -> the JerryScript engine
• Phase II: Data preprocessing, Activity Recognition and Classification
-> MATLAB
-> Python
-> Classifiers parameter tuning -> Python scikit-Learn exhaustive search method GridSearchcv
(with 3-fold-cross-validation and f1-measure scoring)
• Validation method
-> 3-fold-cross-validation
1622
Related works
Introduction
Proposed Method
Evaluation scenarios
Evaluation scenarios
I. Effect of applying proposed method on two data sets (serve & forehand) (Smartwatch ,UTD-MHAD)
II. Proposed method classification results for smartwatch total data set (serve, forehand and
backhand)
III. Classification results & accuracy improvement on UTD-MHAD total data set (27 human
actions)
PLUS
• Effect of Principle Component Analysis(PCA) on classification performance
Introduction
Related works
Proposed Method
Evaluation scenarios
1722
I. Effect of proposed method on two datasets
Classification results for Smartwatch
two strokes dataset:
• Random forest: 1 tree , maxdepth 3
• LSVC: trade off c=0.001
• KNN: 3 neighbours
Classification results for UTD-MHAD two
strokes dataset:
• Random forest: 100 tree , maxdepth 4
• LSVC: trade off c=0.01
• KNN: 1 neighbours
Introduction
Related works
Proposed Method
Evaluation scenarios
1822
II. Classification results of smartwatch data
set
Classification results for three strokes Smartwatch data set:
• Random forest: 15 tree , maxdepth 10
• LSVC: trade off 0.001
• KNN: 3 neighbours
Introduction
Related works
Proposed Method
Evaluation scenarios
1922
III. Classification results
Classification results for 27-actions in UTD-MHAD:
• Random forest: 100 tree , maxdepth 25
• LSVC: trade off 0.01
• KNN: 3 neighbours
Introduction
Related works
Proposed Method
Evaluation scenarios
2022
III. Classification accuracy improvement
Improvement in classification accuracy by more than 30%
Introduction
Related works
Proposed Method
Evaluation scenarios
2122
Conclusion
• We can elevate poor data quality of the smartwatch device with data
preprocessing methods
• Activity recognition with such dataset is possible
• We can utilize these highly resource constraint devices for sport activity
detection
Introduction
Related works
Proposed Method
Evaluation scenarios
2222
Thank You!

Weitere ähnliche Inhalte

Was ist angesagt?

Conectividad inalámbrica para Internet de las cosas(Telecomunicaciones)
Conectividad inalámbrica para Internet de las cosas(Telecomunicaciones)Conectividad inalámbrica para Internet de las cosas(Telecomunicaciones)
Conectividad inalámbrica para Internet de las cosas(Telecomunicaciones)
SANTIAGO PABLO ALBERTO
 
Mark mc granaghan presentation
Mark mc granaghan presentationMark mc granaghan presentation
Mark mc granaghan presentation
Green17Creative
 

Was ist angesagt? (19)

A Unified Semantic Engine for Internet of Things and Smart Cities: From Senso...
A Unified Semantic Engine for Internet of Things and Smart Cities: From Senso...A Unified Semantic Engine for Internet of Things and Smart Cities: From Senso...
A Unified Semantic Engine for Internet of Things and Smart Cities: From Senso...
 
Logical design of io t
Logical design of io tLogical design of io t
Logical design of io t
 
ISWC 2016 Tutorial: Semantic Web of Things M3 framework & FIESTA-IoT EU project
ISWC 2016 Tutorial: Semantic Web of Things  M3 framework & FIESTA-IoT EU projectISWC 2016 Tutorial: Semantic Web of Things  M3 framework & FIESTA-IoT EU project
ISWC 2016 Tutorial: Semantic Web of Things M3 framework & FIESTA-IoT EU project
 
Conectividad inalámbrica para Internet de las cosas(Telecomunicaciones)
Conectividad inalámbrica para Internet de las cosas(Telecomunicaciones)Conectividad inalámbrica para Internet de las cosas(Telecomunicaciones)
Conectividad inalámbrica para Internet de las cosas(Telecomunicaciones)
 
DICYPS
DICYPSDICYPS
DICYPS
 
Digital economy presentation
Digital economy presentationDigital economy presentation
Digital economy presentation
 
IoT Analytics from Edge to Cloud - using IBM Informix
IoT Analytics from Edge to Cloud - using IBM InformixIoT Analytics from Edge to Cloud - using IBM Informix
IoT Analytics from Edge to Cloud - using IBM Informix
 
Vicinity glo tsummit yajuan guan
Vicinity glo tsummit yajuan guanVicinity glo tsummit yajuan guan
Vicinity glo tsummit yajuan guan
 
IoT heap 1
IoT heap 1IoT heap 1
IoT heap 1
 
Green Button Technical Overview
Green Button Technical OverviewGreen Button Technical Overview
Green Button Technical Overview
 
Naming, Search and Discovery in IoT: Issues and proposed solutions in the FP7...
Naming, Search and Discovery in IoT: Issues and proposed solutions in the FP7...Naming, Search and Discovery in IoT: Issues and proposed solutions in the FP7...
Naming, Search and Discovery in IoT: Issues and proposed solutions in the FP7...
 
IoT-A ARM
IoT-A ARMIoT-A ARM
IoT-A ARM
 
IPCA - Intelligent Protocol Content Analysis
IPCA - Intelligent Protocol Content AnalysisIPCA - Intelligent Protocol Content Analysis
IPCA - Intelligent Protocol Content Analysis
 
Lecture 5
Lecture 5Lecture 5
Lecture 5
 
IoT Stream: A Lightweight Ontology for Internet of Things Data Streams (GIoTS...
IoT Stream: A Lightweight Ontology for Internet of Things Data Streams (GIoTS...IoT Stream: A Lightweight Ontology for Internet of Things Data Streams (GIoTS...
IoT Stream: A Lightweight Ontology for Internet of Things Data Streams (GIoTS...
 
Web Techologies and Privacy policies for the Smart Grid
Web Techologies and Privacy policies for the Smart GridWeb Techologies and Privacy policies for the Smart Grid
Web Techologies and Privacy policies for the Smart Grid
 
International Journal of Computational Science and Information Technology (IJ...
International Journal of Computational Science and Information Technology (IJ...International Journal of Computational Science and Information Technology (IJ...
International Journal of Computational Science and Information Technology (IJ...
 
AGILE Gateway for Internet of Things
AGILE Gateway for Internet of ThingsAGILE Gateway for Internet of Things
AGILE Gateway for Internet of Things
 
Mark mc granaghan presentation
Mark mc granaghan presentationMark mc granaghan presentation
Mark mc granaghan presentation
 

Ähnlich wie Tennis stroke detection using inertial data of a smartwatchtion

RS in the context of Big Data-v4
RS in the context of Big Data-v4RS in the context of Big Data-v4
RS in the context of Big Data-v4
Khadija Atiya
 
Data quality evaluation & orbit identification from scatterometer
Data quality evaluation & orbit identification from scatterometerData quality evaluation & orbit identification from scatterometer
Data quality evaluation & orbit identification from scatterometer
Mudit Dholakia
 
Term Paper Presentation
Term Paper PresentationTerm Paper Presentation
Term Paper Presentation
Shubham Singh
 
A Rules-Based Service for Suggesting Visualizations to Analyze Earth Science ...
A Rules-Based Service for Suggesting Visualizations to Analyze Earth Science ...A Rules-Based Service for Suggesting Visualizations to Analyze Earth Science ...
A Rules-Based Service for Suggesting Visualizations to Analyze Earth Science ...
Anirudh Prabhu
 

Ähnlich wie Tennis stroke detection using inertial data of a smartwatchtion (20)

RS in the context of Big Data-v4
RS in the context of Big Data-v4RS in the context of Big Data-v4
RS in the context of Big Data-v4
 
Data quality evaluation & orbit identification from scatterometer
Data quality evaluation & orbit identification from scatterometerData quality evaluation & orbit identification from scatterometer
Data quality evaluation & orbit identification from scatterometer
 
Using bluemix predictive analytics service in Node-RED
Using bluemix predictive analytics service in Node-REDUsing bluemix predictive analytics service in Node-RED
Using bluemix predictive analytics service in Node-RED
 
Optimization of network traffic anomaly detection using machine learning
Optimization of network traffic anomaly detection using machine learning Optimization of network traffic anomaly detection using machine learning
Optimization of network traffic anomaly detection using machine learning
 
Performance Comparison of Dimensionality Reduction Methods using MCDR
Performance Comparison of Dimensionality Reduction Methods using MCDRPerformance Comparison of Dimensionality Reduction Methods using MCDR
Performance Comparison of Dimensionality Reduction Methods using MCDR
 
International Journal of Computational Engineering Research(IJCER)
International Journal of Computational Engineering Research(IJCER)International Journal of Computational Engineering Research(IJCER)
International Journal of Computational Engineering Research(IJCER)
 
Automated Discovery of Performance Regressions in Enterprise Applications
Automated Discovery of Performance Regressions in Enterprise ApplicationsAutomated Discovery of Performance Regressions in Enterprise Applications
Automated Discovery of Performance Regressions in Enterprise Applications
 
DWDM syllabus.doc
DWDM syllabus.docDWDM syllabus.doc
DWDM syllabus.doc
 
Artificial Intelligence for Automating Data Analysis
Artificial Intelligence for Automating Data AnalysisArtificial Intelligence for Automating Data Analysis
Artificial Intelligence for Automating Data Analysis
 
M3AT: Monitoring Agents Assignment Model for the Data-Intensive Applications
M3AT: Monitoring Agents Assignment Model for the Data-Intensive ApplicationsM3AT: Monitoring Agents Assignment Model for the Data-Intensive Applications
M3AT: Monitoring Agents Assignment Model for the Data-Intensive Applications
 
Fi cloudpresentationgyrardaugust2015 v2
Fi cloudpresentationgyrardaugust2015 v2Fi cloudpresentationgyrardaugust2015 v2
Fi cloudpresentationgyrardaugust2015 v2
 
GIoTS 2022 YuukiTakagi
GIoTS 2022 YuukiTakagiGIoTS 2022 YuukiTakagi
GIoTS 2022 YuukiTakagi
 
8th TUC Meeting - Tim Hegeman (TU Delft). Social Network Benchmark, Analytics...
8th TUC Meeting - Tim Hegeman (TU Delft). Social Network Benchmark, Analytics...8th TUC Meeting - Tim Hegeman (TU Delft). Social Network Benchmark, Analytics...
8th TUC Meeting - Tim Hegeman (TU Delft). Social Network Benchmark, Analytics...
 
Big Data and IOT
Big Data and IOTBig Data and IOT
Big Data and IOT
 
Term Paper Presentation
Term Paper PresentationTerm Paper Presentation
Term Paper Presentation
 
Test Tool for Industrial Ethernet Network Performance (June 2009)
Test Tool for Industrial Ethernet Network Performance (June 2009)Test Tool for Industrial Ethernet Network Performance (June 2009)
Test Tool for Industrial Ethernet Network Performance (June 2009)
 
Iiwas19 yamazaki slide
Iiwas19 yamazaki slideIiwas19 yamazaki slide
Iiwas19 yamazaki slide
 
A Rules-Based Service for Suggesting Visualizations to Analyze Earth Science ...
A Rules-Based Service for Suggesting Visualizations to Analyze Earth Science ...A Rules-Based Service for Suggesting Visualizations to Analyze Earth Science ...
A Rules-Based Service for Suggesting Visualizations to Analyze Earth Science ...
 
Ds for finance day 3
Ds for finance day 3Ds for finance day 3
Ds for finance day 3
 
Intelligent Career Guidance System.pptx
Intelligent Career Guidance System.pptxIntelligent Career Guidance System.pptx
Intelligent Career Guidance System.pptx
 

Mehr von Maynooth University

A two-dimensional self-coordination mechanism of agents in a minority game
A two-dimensional self-coordination mechanism of agents in a minority gameA two-dimensional self-coordination mechanism of agents in a minority game
A two-dimensional self-coordination mechanism of agents in a minority game
Maynooth University
 
PAMS: A new position-aware multi-sensor dataset for human activity recognitio...
PAMS: A new position-aware multi-sensor dataset for human activity recognitio...PAMS: A new position-aware multi-sensor dataset for human activity recognitio...
PAMS: A new position-aware multi-sensor dataset for human activity recognitio...
Maynooth University
 
Delaunay based two-phase algorithm for connected cover in WSNs
Delaunay based two-phase algorithm for connected cover in WSNsDelaunay based two-phase algorithm for connected cover in WSNs
Delaunay based two-phase algorithm for connected cover in WSNs
Maynooth University
 

Mehr von Maynooth University (8)

Evidential fine-grained event localization using Twitter
Evidential fine-grained event localization using TwitterEvidential fine-grained event localization using Twitter
Evidential fine-grained event localization using Twitter
 
A two-dimensional self-coordination mechanism of agents in a minority game
A two-dimensional self-coordination mechanism of agents in a minority gameA two-dimensional self-coordination mechanism of agents in a minority game
A two-dimensional self-coordination mechanism of agents in a minority game
 
PAMS: A new position-aware multi-sensor dataset for human activity recognitio...
PAMS: A new position-aware multi-sensor dataset for human activity recognitio...PAMS: A new position-aware multi-sensor dataset for human activity recognitio...
PAMS: A new position-aware multi-sensor dataset for human activity recognitio...
 
Delaunay based two-phase algorithm for connected cover in WSNs
Delaunay based two-phase algorithm for connected cover in WSNsDelaunay based two-phase algorithm for connected cover in WSNs
Delaunay based two-phase algorithm for connected cover in WSNs
 
Quality of Claim Metrics in Social Sensing Systems: A case study on IranDeal
Quality of Claim Metrics in Social Sensing Systems: A case study on IranDealQuality of Claim Metrics in Social Sensing Systems: A case study on IranDeal
Quality of Claim Metrics in Social Sensing Systems: A case study on IranDeal
 
NoSQL Data Architecture Patterns
NoSQL Data ArchitecturePatternsNoSQL Data ArchitecturePatterns
NoSQL Data Architecture Patterns
 
NoSQL Consepts
NoSQL ConseptsNoSQL Consepts
NoSQL Consepts
 
Chapter1: NoSQL: It’s about making intelligent choices
Chapter1: NoSQL: It’s about making intelligent choicesChapter1: NoSQL: It’s about making intelligent choices
Chapter1: NoSQL: It’s about making intelligent choices
 

Kürzlich hochgeladen

➥🔝 7737669865 🔝▻ mahisagar Call-girls in Women Seeking Men 🔝mahisagar🔝 Esc...
➥🔝 7737669865 🔝▻ mahisagar Call-girls in Women Seeking Men  🔝mahisagar🔝   Esc...➥🔝 7737669865 🔝▻ mahisagar Call-girls in Women Seeking Men  🔝mahisagar🔝   Esc...
➥🔝 7737669865 🔝▻ mahisagar Call-girls in Women Seeking Men 🔝mahisagar🔝 Esc...
amitlee9823
 
Call Girls Bannerghatta Road Just Call 👗 7737669865 👗 Top Class Call Girl Ser...
Call Girls Bannerghatta Road Just Call 👗 7737669865 👗 Top Class Call Girl Ser...Call Girls Bannerghatta Road Just Call 👗 7737669865 👗 Top Class Call Girl Ser...
Call Girls Bannerghatta Road Just Call 👗 7737669865 👗 Top Class Call Girl Ser...
amitlee9823
 
Call Girls In Doddaballapur Road ☎ 7737669865 🥵 Book Your One night Stand
Call Girls In Doddaballapur Road ☎ 7737669865 🥵 Book Your One night StandCall Girls In Doddaballapur Road ☎ 7737669865 🥵 Book Your One night Stand
Call Girls In Doddaballapur Road ☎ 7737669865 🥵 Book Your One night Stand
amitlee9823
 
👉 Amritsar Call Girl 👉📞 6367187148 👉📞 Just📲 Call Ruhi Call Girl Phone No Amri...
👉 Amritsar Call Girl 👉📞 6367187148 👉📞 Just📲 Call Ruhi Call Girl Phone No Amri...👉 Amritsar Call Girl 👉📞 6367187148 👉📞 Just📲 Call Ruhi Call Girl Phone No Amri...
👉 Amritsar Call Girl 👉📞 6367187148 👉📞 Just📲 Call Ruhi Call Girl Phone No Amri...
karishmasinghjnh
 
➥🔝 7737669865 🔝▻ malwa Call-girls in Women Seeking Men 🔝malwa🔝 Escorts Ser...
➥🔝 7737669865 🔝▻ malwa Call-girls in Women Seeking Men  🔝malwa🔝   Escorts Ser...➥🔝 7737669865 🔝▻ malwa Call-girls in Women Seeking Men  🔝malwa🔝   Escorts Ser...
➥🔝 7737669865 🔝▻ malwa Call-girls in Women Seeking Men 🔝malwa🔝 Escorts Ser...
amitlee9823
 
Junnasandra Call Girls: 🍓 7737669865 🍓 High Profile Model Escorts | Bangalore...
Junnasandra Call Girls: 🍓 7737669865 🍓 High Profile Model Escorts | Bangalore...Junnasandra Call Girls: 🍓 7737669865 🍓 High Profile Model Escorts | Bangalore...
Junnasandra Call Girls: 🍓 7737669865 🍓 High Profile Model Escorts | Bangalore...
amitlee9823
 
Call Girls In Attibele ☎ 7737669865 🥵 Book Your One night Stand
Call Girls In Attibele ☎ 7737669865 🥵 Book Your One night StandCall Girls In Attibele ☎ 7737669865 🥵 Book Your One night Stand
Call Girls In Attibele ☎ 7737669865 🥵 Book Your One night Stand
amitlee9823
 
Call Girls Jalahalli Just Call 👗 7737669865 👗 Top Class Call Girl Service Ban...
Call Girls Jalahalli Just Call 👗 7737669865 👗 Top Class Call Girl Service Ban...Call Girls Jalahalli Just Call 👗 7737669865 👗 Top Class Call Girl Service Ban...
Call Girls Jalahalli Just Call 👗 7737669865 👗 Top Class Call Girl Service Ban...
amitlee9823
 
Probability Grade 10 Third Quarter Lessons
Probability Grade 10 Third Quarter LessonsProbability Grade 10 Third Quarter Lessons
Probability Grade 10 Third Quarter Lessons
JoseMangaJr1
 
Abortion pills in Jeddah | +966572737505 | Get Cytotec
Abortion pills in Jeddah | +966572737505 | Get CytotecAbortion pills in Jeddah | +966572737505 | Get Cytotec
Abortion pills in Jeddah | +966572737505 | Get Cytotec
Abortion pills in Riyadh +966572737505 get cytotec
 
Vip Mumbai Call Girls Thane West Call On 9920725232 With Body to body massage...
Vip Mumbai Call Girls Thane West Call On 9920725232 With Body to body massage...Vip Mumbai Call Girls Thane West Call On 9920725232 With Body to body massage...
Vip Mumbai Call Girls Thane West Call On 9920725232 With Body to body massage...
amitlee9823
 

Kürzlich hochgeladen (20)

➥🔝 7737669865 🔝▻ mahisagar Call-girls in Women Seeking Men 🔝mahisagar🔝 Esc...
➥🔝 7737669865 🔝▻ mahisagar Call-girls in Women Seeking Men  🔝mahisagar🔝   Esc...➥🔝 7737669865 🔝▻ mahisagar Call-girls in Women Seeking Men  🔝mahisagar🔝   Esc...
➥🔝 7737669865 🔝▻ mahisagar Call-girls in Women Seeking Men 🔝mahisagar🔝 Esc...
 
April 2024 - Crypto Market Report's Analysis
April 2024 - Crypto Market Report's AnalysisApril 2024 - Crypto Market Report's Analysis
April 2024 - Crypto Market Report's Analysis
 
Call Girls Bannerghatta Road Just Call 👗 7737669865 👗 Top Class Call Girl Ser...
Call Girls Bannerghatta Road Just Call 👗 7737669865 👗 Top Class Call Girl Ser...Call Girls Bannerghatta Road Just Call 👗 7737669865 👗 Top Class Call Girl Ser...
Call Girls Bannerghatta Road Just Call 👗 7737669865 👗 Top Class Call Girl Ser...
 
Call Girls In Doddaballapur Road ☎ 7737669865 🥵 Book Your One night Stand
Call Girls In Doddaballapur Road ☎ 7737669865 🥵 Book Your One night StandCall Girls In Doddaballapur Road ☎ 7737669865 🥵 Book Your One night Stand
Call Girls In Doddaballapur Road ☎ 7737669865 🥵 Book Your One night Stand
 
5CL-ADBA,5cladba, Chinese supplier, safety is guaranteed
5CL-ADBA,5cladba, Chinese supplier, safety is guaranteed5CL-ADBA,5cladba, Chinese supplier, safety is guaranteed
5CL-ADBA,5cladba, Chinese supplier, safety is guaranteed
 
👉 Amritsar Call Girl 👉📞 6367187148 👉📞 Just📲 Call Ruhi Call Girl Phone No Amri...
👉 Amritsar Call Girl 👉📞 6367187148 👉📞 Just📲 Call Ruhi Call Girl Phone No Amri...👉 Amritsar Call Girl 👉📞 6367187148 👉📞 Just📲 Call Ruhi Call Girl Phone No Amri...
👉 Amritsar Call Girl 👉📞 6367187148 👉📞 Just📲 Call Ruhi Call Girl Phone No Amri...
 
➥🔝 7737669865 🔝▻ malwa Call-girls in Women Seeking Men 🔝malwa🔝 Escorts Ser...
➥🔝 7737669865 🔝▻ malwa Call-girls in Women Seeking Men  🔝malwa🔝   Escorts Ser...➥🔝 7737669865 🔝▻ malwa Call-girls in Women Seeking Men  🔝malwa🔝   Escorts Ser...
➥🔝 7737669865 🔝▻ malwa Call-girls in Women Seeking Men 🔝malwa🔝 Escorts Ser...
 
Junnasandra Call Girls: 🍓 7737669865 🍓 High Profile Model Escorts | Bangalore...
Junnasandra Call Girls: 🍓 7737669865 🍓 High Profile Model Escorts | Bangalore...Junnasandra Call Girls: 🍓 7737669865 🍓 High Profile Model Escorts | Bangalore...
Junnasandra Call Girls: 🍓 7737669865 🍓 High Profile Model Escorts | Bangalore...
 
Call Girls In Attibele ☎ 7737669865 🥵 Book Your One night Stand
Call Girls In Attibele ☎ 7737669865 🥵 Book Your One night StandCall Girls In Attibele ☎ 7737669865 🥵 Book Your One night Stand
Call Girls In Attibele ☎ 7737669865 🥵 Book Your One night Stand
 
Call Girls Jalahalli Just Call 👗 7737669865 👗 Top Class Call Girl Service Ban...
Call Girls Jalahalli Just Call 👗 7737669865 👗 Top Class Call Girl Service Ban...Call Girls Jalahalli Just Call 👗 7737669865 👗 Top Class Call Girl Service Ban...
Call Girls Jalahalli Just Call 👗 7737669865 👗 Top Class Call Girl Service Ban...
 
Capstone Project on IBM Data Analytics Program
Capstone Project on IBM Data Analytics ProgramCapstone Project on IBM Data Analytics Program
Capstone Project on IBM Data Analytics Program
 
Thane Call Girls 7091864438 Call Girls in Thane Escort service book now -
Thane Call Girls 7091864438 Call Girls in Thane Escort service book now -Thane Call Girls 7091864438 Call Girls in Thane Escort service book now -
Thane Call Girls 7091864438 Call Girls in Thane Escort service book now -
 
Probability Grade 10 Third Quarter Lessons
Probability Grade 10 Third Quarter LessonsProbability Grade 10 Third Quarter Lessons
Probability Grade 10 Third Quarter Lessons
 
VIP Model Call Girls Hinjewadi ( Pune ) Call ON 8005736733 Starting From 5K t...
VIP Model Call Girls Hinjewadi ( Pune ) Call ON 8005736733 Starting From 5K t...VIP Model Call Girls Hinjewadi ( Pune ) Call ON 8005736733 Starting From 5K t...
VIP Model Call Girls Hinjewadi ( Pune ) Call ON 8005736733 Starting From 5K t...
 
Predicting Loan Approval: A Data Science Project
Predicting Loan Approval: A Data Science ProjectPredicting Loan Approval: A Data Science Project
Predicting Loan Approval: A Data Science Project
 
Abortion pills in Jeddah | +966572737505 | Get Cytotec
Abortion pills in Jeddah | +966572737505 | Get CytotecAbortion pills in Jeddah | +966572737505 | Get Cytotec
Abortion pills in Jeddah | +966572737505 | Get Cytotec
 
Midocean dropshipping via API with DroFx
Midocean dropshipping via API with DroFxMidocean dropshipping via API with DroFx
Midocean dropshipping via API with DroFx
 
Vip Mumbai Call Girls Thane West Call On 9920725232 With Body to body massage...
Vip Mumbai Call Girls Thane West Call On 9920725232 With Body to body massage...Vip Mumbai Call Girls Thane West Call On 9920725232 With Body to body massage...
Vip Mumbai Call Girls Thane West Call On 9920725232 With Body to body massage...
 
Call me @ 9892124323 Cheap Rate Call Girls in Vashi with Real Photo 100% Secure
Call me @ 9892124323  Cheap Rate Call Girls in Vashi with Real Photo 100% SecureCall me @ 9892124323  Cheap Rate Call Girls in Vashi with Real Photo 100% Secure
Call me @ 9892124323 Cheap Rate Call Girls in Vashi with Real Photo 100% Secure
 
hybrid Seed Production In Chilli & Capsicum.pptx
hybrid Seed Production In Chilli & Capsicum.pptxhybrid Seed Production In Chilli & Capsicum.pptx
hybrid Seed Production In Chilli & Capsicum.pptx
 

Tennis stroke detection using inertial data of a smartwatchtion

  • 1. 9th International Conference on Computer and Knowledge Engineering (ICCKE 2019), October 24-25 2019, Ferdowsi University of Mashhad Tennis stroke detection using inertial data of a smartwatch Sara Taghavi, Fardjad Davari, Hadi Tabatabaee Malazi, Ahmad Ali Abin Faculty of Computer Science and Engineering, G.C. Shahid Beheshti University, Tehran, Iran
  • 2. Motivation • Analyzing sports activities are changing: • Question is: How accurate a commercial smartwatch can detect tennis strokes? 122
  • 3. Outline: • Introduction • Related works • Proposed method • Evaluation scenarios: i. Effect of applying proposed method on two datasets ii. Proposed method classification results for smartwatch dataset iii. Classification results & classification accuracy improvement Introduction Related works Proposed Method Evaluation scenarios 222
  • 4. Introduction • Smartwatch for Tennis Activity Detection -> its popularity • As it is not an open eco-system -> Challenges: 1. Fluctuating sampling rate of sensor data 2. Limited data transmission rate 3. Highly resourced constraint regarding memory space 4. The processing power adds more constraints Introduction Related works Proposed Method Evaluation scenarios 322
  • 5. Network point of view Introduction Related works Proposed Method Evaluation scenarios 422
  • 6. Related works • M. Kos et al. introduced a system: designed a lightweight device can be worn on the wrist communicates with the personal computer (PC) through USB • Pei et al. in developed a product: Has three modules : sensor, controller and transmission is embedded in tennis racket the motion information is sent to the mobile phone • Dhnesh et al. introduced a platform: Wireless measurement sensors attached to the racket and player’s body work in conjunction with software analysis modules on PC Introduction Related works Proposed Method Evaluation scenarios 522
  • 7. Related works • Previous works in Tennis Activity Detection -> custom-built wearables Advantages: -> Programmable micro-controllers -> Capable of adjusting sufficient memory space An open customizable eco-system High-quality datasets Very few of them are publically available! Introduction Related works Proposed Method Evaluation scenarios 622
  • 8. Proposed Method • Phase one -> The Data Collector Application on the smartwatch • Phase two -> Data pre-processing and Tennis Activity Detection DATA TAD Pre-processed Data ML methods Introduction Related works Proposed Method Evaluation scenarios 722
  • 9. Phase I: Data Collector Application 1. Haptic feedback 2. Logger 3. Sensors helper 4. Messaging helper 5. Client SDK Introduction Related works Proposed Method Evaluation scenarios 822 Smartwatch application is designed with 5 modules:
  • 10. Phase I: Data Collector Application The most important modules are: 1. Sensor helper 2. Messaging helper 3. Client SDK 922 Introduction Related works Proposed Method Evaluation scenarios
  • 11. Data Collection Procedure 1022 Introduction Related works Proposed Method Evaluation scenarios
  • 12. Data sets characteristics • Smartwatch data set: -> Sensors: 3D Accelerometer & 3D Gyroscope(angular velocity) -> Sampling rate: 50Hz -> 8 subjects, 4 try for each stroke -> Strokes : Serve, Forehand, Backhand Same as: • UTD-MHAD multi-modal data set: -> Strokes: Serve , Forehand -> Custom-built wearable inertial unit -> Publicly available Related works Introduction Proposed Method Evaluation scenarios 1122 Introduction Proposed Method Evaluation scenarios
  • 13. Phase II: Data pre-processing & Activity Recognition -> To elevate data quality -> Reduce effects of poor data quality for AR process Steps Include: • Data Cleansing -> exclude irrelevant data • Stream Alignment -> equal number of windows containing data points • Signal Processing -> to reduce noise -> high-pass and low-pass filters Related works Introduction Proposed Method Evaluation scenarios 1222 Introduction Proposed Method Evaluation scenarios
  • 14. Phase II: Smartwatch Signal Processing Related works Introduction Proposed Method Evaluation scenarios 1322 Related works Introduction Proposed Method Evaluation scenarios
  • 15. Phase II: UTD-MHAD Signal Processing Related works Introduction Proposed Method Evaluation scenarios 1422 Related works Introduction Proposed Method Evaluation scenarios
  • 16. Phase II: Feature extraction & Feature importance • An 18 dimension data matrix for each data steam • Segmentation method: Fix time-based sliding window with overlap Related works Introduction Proposed Method Evaluation scenarios 1522 Related works Introduction Proposed Method Evaluation scenarios Features: Feature importance:
  • 17. Tools • Smart watch device : Fitbit Ionic™ Watch • Phase I: Data Collector Application development -> IDE: Fitbit online Studio -> JavaScript -> compile, bundle and optimize -> TypeScript compiler and rollup.js -> JavaScript is run on the device using -> the JerryScript engine • Phase II: Data preprocessing, Activity Recognition and Classification -> MATLAB -> Python -> Classifiers parameter tuning -> Python scikit-Learn exhaustive search method GridSearchcv (with 3-fold-cross-validation and f1-measure scoring) • Validation method -> 3-fold-cross-validation 1622 Related works Introduction Proposed Method Evaluation scenarios
  • 18. Evaluation scenarios I. Effect of applying proposed method on two data sets (serve & forehand) (Smartwatch ,UTD-MHAD) II. Proposed method classification results for smartwatch total data set (serve, forehand and backhand) III. Classification results & accuracy improvement on UTD-MHAD total data set (27 human actions) PLUS • Effect of Principle Component Analysis(PCA) on classification performance Introduction Related works Proposed Method Evaluation scenarios 1722
  • 19. I. Effect of proposed method on two datasets Classification results for Smartwatch two strokes dataset: • Random forest: 1 tree , maxdepth 3 • LSVC: trade off c=0.001 • KNN: 3 neighbours Classification results for UTD-MHAD two strokes dataset: • Random forest: 100 tree , maxdepth 4 • LSVC: trade off c=0.01 • KNN: 1 neighbours Introduction Related works Proposed Method Evaluation scenarios 1822
  • 20. II. Classification results of smartwatch data set Classification results for three strokes Smartwatch data set: • Random forest: 15 tree , maxdepth 10 • LSVC: trade off 0.001 • KNN: 3 neighbours Introduction Related works Proposed Method Evaluation scenarios 1922
  • 21. III. Classification results Classification results for 27-actions in UTD-MHAD: • Random forest: 100 tree , maxdepth 25 • LSVC: trade off 0.01 • KNN: 3 neighbours Introduction Related works Proposed Method Evaluation scenarios 2022
  • 22. III. Classification accuracy improvement Improvement in classification accuracy by more than 30% Introduction Related works Proposed Method Evaluation scenarios 2122
  • 23. Conclusion • We can elevate poor data quality of the smartwatch device with data preprocessing methods • Activity recognition with such dataset is possible • We can utilize these highly resource constraint devices for sport activity detection Introduction Related works Proposed Method Evaluation scenarios 2222