Activity Recognition Classification in Action

Slides:



Advertisements
Similar presentations
Outline Activity recognition applications
Advertisements

Data Mining Methodology 1. Why have a Methodology  Don’t want to learn things that aren’t true May not represent any underlying reality ○ Spurious correlation.
Gary M. Weiss and Jeffrey Lockhart Fordham University, New York, NY 1UbiMI UBICOMP Sept
Smartphone-based Activity Recognition for Pervasive Healthcare - Utilizing Cloud Infrastructure for Data Modeling Bingchuan Yuan, John Herbert University.
Indian Statistical Institute Kolkata
EHealth Workshop 2003Virginia Tech e-Textiles Group An E-Textile System for Motion Analysis Mark Jones, Thurmon Lockhart, and Thomas Martin Virginia Tech.
Chapter 7 – Classification and Regression Trees
Chapter 7 – Classification and Regression Trees
Studying Relationships between Human Posture and Health Risk Factors during Sedentary Activities Tejas Srinivasan Mentors: Vladimir Pavlovic Saehoon Yi.
A Practical Approach to Recognizing Physical Activities Jonathan Lester, Tanzeem Choudhury, and Gaetano Borriello In Proceedings of the Fourth International.
Activity Recognition from User- Annotated Acceleration Data Presented by James Reinebold CSCI 546.
Learning from Multiple Outlooks Maayan Harel and Shie Mannor ICML 2011 Presented by Minhua Chen.
Gary M. Weiss Fordham University
Feature Extraction Spring Semester, Accelerometer Based Gestural Control of Browser Applications M. Kauppila et al., In Proc. of Int. Workshop on.
南台科技大學 資訊工程系 Posture Monitoring System for Context Awareness in Mobile Computing Authors: Jonghun Baek and Byoung-Ju Yun Adviser: Yu-Chiang Li Speaker:
July 25, 2010 SensorKDD Activity Recognition Using Cell Phone Accelerometers Jennifer Kwapisz, Gary Weiss, Samuel Moore Department of Computer &
September Activity Recognition and Biometric Identification Using Cell Phone Accelerometers WISDM Project Department of Computer & Info. Science.
TEMPLATE DESIGN © Detecting User Activities Using the Accelerometer on Android Smartphones Sauvik Das, Supervisor: Adrian.
Chapter 9 – Classification and Regression Trees
Slice&Dice: recognizing food preparation activities using embedded accelerometers Cuong Pham & Patrick Olivier Culture Lab School of Computing Science.
Human Activity Recognition Using Accelerometer on Smartphones
Department of Computer and Electrical Engineering A Study of Time-based Features and Regularity of Manipulation to Improve the Detection of Eating Activity.
Abstract: Accelerometers As part of the smartphone philosophy, every phone has a wide variety of sensors available to the user. Sensors include light and.
1 CS 391L: Machine Learning: Experimental Evaluation Raymond J. Mooney University of Texas at Austin.
Outdoors Indoors For your Health Ends in a “K’ Mixed.
1 Research Question  Can a vision-based mobile robot  with limited computation and memory,  and rapidly varying camera positions,  operate autonomously.
J.-Y. Yang, J.-S. Wang and Y.-P. Chena, Using acceleration measurements for activity recognition: An effective learning algorithm for constructing neural.
How Good is a Model? How much information does AIC give us? –Model 1: 3124 –Model 2: 2932 –Model 3: 2968 –Model 4: 3204 –Model 5: 5436.
Saisakul Chernbumroong, Shuang Cang, Anthony Atkins, Hongnian Yu Expert Systems with Applications 40 (2013) 1662–1674 Elderly activities recognition and.
An E-Textiles. Virginia Tech e-Textiles Group Design of an e-textile computer architecture – Networking – Fault tolerance – Power aware – Programming.
A Behavioral Biometrics User Authentication Study Using Android Device Accelerometer and Gyroscope Data Jonathan Lee, Aliza Levinger, Beqir Simnica, Khushbu.
Work supported by NSF Grant No and numerous Fordham University grants
Mobile Activity Recognition
Data Mining ICCM
Evaluating Classifiers
Telepath: Sensory Offloading for Wearable Devices
Where is it? Find the Action.
My Tiny Ping-Pong Helper
Practice & Communication of Science From Distributions to Confidence
Tracking Mobile Web Users Through Motion Sensors: Attacks and Defenses
How Good is a Model? How much information does AIC give us?
Posture Monitoring System for Context Awareness in Mobile Computing
Transport mode detection in the city of Lyon using mobile phone sensors Jorge Chong Internship for MLDM M1 Jean Monnet University
BAYESIAN APPROACH FOR INDOOR HUMAN ACTIVITY MONITORING
Walking Speed Detection from 5G Prototype System
Machine Learning for the Quantified Self
Recognizing Smoking Gestures with Inertial Measurements Unit (IMU)
Vijay Srinivasan Thomas Phan
Map My Ride Application
Transportation Mode Recognition using Smartphone Sensor Data
From Distributions to Confidence
Massachusetts Institute of Technology
Human Activity Recognition Using Smartphone Sensor Data
Mobile Sensor-Based Biometrics Using Common Daily Activities
Chao Xu, Parth H. Pathak, et al. HotMobile’15
Anindya Maiti, Murtuza Jadliwala, Jibo He Igor Bilogrevic
iSRD Spam Review Detection with Imbalanced Data Distributions
WISDM Activity Recognition & Biometrics Applications of Classification
Approaching an ML Problem
Xin Qi, Matthew Keally, Gang Zhou, Yantao Li, Zhen Ren
Sensor Fusion Localization and Navigation for Visually Impaired People
Data Transformations targeted at minimizing experimental variance
Ensemble learning Reminder - Bagging of Trees Random Forest
William Fadel, Ph.D. August 1, 2018
Visual Recognition of American Sign Language Using Hidden Markov Models 문현구 문현구.
MAS 622J Course Project Classification of Affective States - GP Semi-Supervised Learning, SVM and kNN Hyungil Ahn
Raveen Wijewickrama Anindya Maiti Murtuza Jadliwala
Machine Learning: Lecture 5
Mole: Motion Leaks through Smartwatch Sensors
Presenter: Donovan Orn
Presentation transcript:

Activity Recognition Classification in Action

Mobile Activity Recognition Mobile devices like smartphones and smartwatches have many sensors Some sensors measure motion Tri-axial accelerometer Gyroscope GPS and other location sensors Activity Recognition is now pretty common but wasn’t when this research started It took a year for our Fitbit order to be delivered

What is Activity Recognition? Identifying a user’s activity based on data In our case the mobile sensor data from the accelerometer and gyroscope What type of data mining task is this? Classification How would you formulate this as a classification task? Not so obvious if you have not read the paper, since time dimension complicates things

More on Activity Recognition Examples of activities Walking, jogging, running, jumping, washing dishes, playing basketball, reading, partying, studying, eating, drinking, etc Why do we care? Context sensitive “smart” devices Fitness and health applications To track what we do for other purposes

The Data The data is collected at 20 Hz A timestamped sequence of numbers for each of 3 dimensions for both sensors

Walking Data Watch Gyroscope Phone accelerometer

Phone Accelerometer (Jogging)

Phone Accelerometer (Standing)

WISDM Activity Recognition Studies 2010 study using only smartphones Good results, but only 6 basic activities (29 subjects) More refined studies over next few years, including impact of personal models 2017 study 18 activities and 51 test subjects Includes eating actvities Sensors Evaluates accel and gyro on watch and phone (4 sensors) Evaluates 5 fused sensors

The 2016 Smartwatch Activities General Activities Walking* Jogging* Climbing Stairs* Sitting* Standing* Kicking Soccer Ball General Activities (hand-oriented) Dribbling Basketball Playing Catch with Tennis Ball Typing Handwriting Clapping Brushing Teeth Folding Clothes Eating Activities (hand-oriented) Eating Pasta Eating Soup Eating Sandwich Eating Chips Drinking from a Cup * These used in the 2010 smartphone study

Formulation as Classification Take the raw time series sensor data for non- overlapping 10 second chunks and create one example Use higher level features to describe behavior over the 10 second period This is data transformation Mapping the data to a very different representation This is because most classification algorithms assume examples, not time series data

High Level Features: 43 Total Average[3]: Average acceleration (per axis) Standard Deviation[3]: SD per axis Average Absolute difference[3]: per axis Average Resultant Acceleration[3]: average of square root of sum of squares of 3 values Time Between Peaks[3] Binned Distribution[30]: For each axis take max – min value, create 10 equal sized bins, and record fraction in each bin

Types of Models Impersonal Models Personal Models Generated using data from a panel of other users Build model based on 50 subjects and test on 51st Repeat 51 times so evaluate on every subject using all other subjects Personal Models Generated using data from the intended user. Must generate 51 models and carefully partition data for each subject

Results

2010 Study using Impersonal Model (IB3 Method) 72.4% Accuracy  Predicted Class Walking Jogging Stairs Sitting Standing Lying Down Actual Class 2209 46 789 2 4 45 1656 148 1 412 54 869 3 10 47 553 30 241 8 57 6 448 5 7 301 13 131

2010 Study using Personal Model (IB3 Method) 98.4% accuracy  Predicted Class Walking Jogging Stairs Sitting Standing Lying Down Actual Class 3033 1 24 4 1788 42 1292 870 2 6 5 11 509 8 7 442

2010 Study Accuracy Results % of Records Correctly Classified Personal Universal Straw Man IB3 J48 NN Walking 99.2 97.5 99.1 72.4 77.3 60.6 37.7 Jogging 99.6 98.9 99.9 89.5 89.7 89.9 22.8 Stairs 96.5 91.7 98.0 64.9 56.7 67.6 16.5 Sitting 98.6 97.6 97.7 62.8 78.0 10.9 Standing 96.8 96.4 97.3 85.8 92.0 93.6 6.4 Lying Down 95.9 95.0 96.9 28.6 26.2 60.7 5.7 Overall 98.4 96.6 98.7 74.9 71.2

Personal Model Accuracy (RF)

Impersonal Model Accuracy (RF)

Accuracy of Different Classification Algorithms

Personal Model Learning Curves The x-axis represents the amount of training data per activity

Impersonal Model Learning Curves The x-axis represents the amount of training data per activity per panelist (50 panelists)

Personal Model Learning Curves (RF with varying sensors)

Impersonal Model Learning Curves (RF with varying sensors)

Impersonal Model Learning Curves (varying number of panelists)

Hybrid Models Much related work uses hybrid models rather than personal and impersonal models Hybrid models are treated as if they are impersonal models, but allow the panelist to also be in the test set Easy to generate since can use cross-validation on labeled data with many users But this is cheating! One might assume that small overlap is not a problem since if 50 people in the panel, then each test subject has only a 2% overlap with training data Our prior research shows that it is a huge problem and hybrid models perform like personal models The classifier must implicitly identify the subject, and our work on biometrics shows that this is achievable

Actitracker The phone-based research was incorporated into a deployed app/system called Actitracker The development effort to handle real-time activity recognition was substantial Actitracker is no longer supported