Transcript Sensors - Fordham University

Work supported by NSF Grant No. 1116124
and numerous Fordham University grants
Gary M. Weiss
Chair, Dept. of Computer & Info Science
Fordham University
[email protected]
storm.cis.fordham.edu/~gweiss

A smart phone is a ___________
(think about separate devices it can replace)
 Mobile “phone”
 Internet connected computer (web, email, etc.)
 Music device (MP3 player)
 Gaming device
 Camera & video recorder
 PDA: calendar, address book, etc.
 GPS-enabled map and guide
 Sensor array
7/25/2016
Gary M. Weiss
DMIN Invited Talk
2


A watch
More convenient interface to your phone
 For text, email, etc.


Not a heck of a lot more at the moment
Sensor array
 Many of the same sensors as your phone
 Useful for health and fitness applications
7/25/2016
Gary M. Weiss
DMIN Invited Talk
3

What sensors are found on smartphones & watches?










Audio sensor (microphone) [not all watches]
Image sensor (camera, video recorder) [few watches]
Location sensor (GPS, cell tower, WiFi) [phone only]
Proximity and motion sensor (infrared) [phone]
Light sensor
Tri-Axial Accelerometer; Gyroscope [most watches]
Magnetic field sensor/compass
Temperature and humidity sensor
Pressure sensor (barometer)
Heart Rate Sensor [some watches few phones*]
* virtual sensor usually reads color of finger
7/25/2016
Gary M. Weiss
DMIN Invited Talk
4

What might be some advantages of using watch vs
phone?
 Worn in consistent on-body position
▪ Phones are shifted a lot
▪ Women tend not to “wear” phones– rarely reside in front pants
pocket
 Wrist position superior for tracking hand-based
activities
▪ We rely on hands for most tasks-- quite useful if we want to go
beyond just fitness
▪ Eating, typing, writing etc.
7/25/2016
Gary M. Weiss
DMIN Invited Talk
5

Data mining: application of computational methods
to extract knowledge from data
 Most data mining involves inferring predictive models,
often for classification

Sensor mining: application of computational
methods to extract knowledge from sensor data
7/25/2016
Gary M. Weiss
DMIN Invited Talk
6
Provide basic knowledge of this area
 Encourage/motivate/promote R&D

 Interesting work remains to be done
 Good intro to Internet of Things
▪ I am transitioning now to working with cheap sensors
Name:
SensorTag
Battery Life: Approximately 1 year
User Interface: Via smartphone app
Sensors:
Light, Temperature, Accelerometer
Gyroscope, Magnetometer, Pressure
Humidity, Microphone, Magnetic Sensor
Dimensions: Around 2" x 2“
Cost:
$20
7/25/2016
Gary M. Weiss
DMIN Invited Talk
7




Sensor Overview
Activity Recognition
Biometrics
Future when sensors are ubiquitous
7/25/2016
Gary M. Weiss
DMIN Invited Talk
8
7/25/2016
Gary M. Weiss
DMIN Invited Talk
9

Tri-axial accelerometer
 Present in virtually all smart phones and smart watches
 On Android & iOS default range is +2g to -2g
▪ Axes are fixed relative to phone and hence changes as phone shifts
 Sampling rates 20-50 Hz
▪ Study found 20Hz required for activity recognition4
▪ We found could not reliably sample beyond 20Hz18
 Uses:
▪ originally mainly for game play and shifting display orientation
▪ Now used for activity recognition & fitness apps
▪ We use for biometrics
7/25/2016
Gary M. Weiss
DMIN Invited Talk
10

Gyroscope
 Determines position and orientation of device
 Measures rotation in radians/sec about each axis
 Sensitivity to rotation is more robust to motion than
accelerometer

Uses
 Compasses and navigation
 Recognition of spatial motions and gestures
 We use for activity recognition and biometrics
▪ Not as useful as accelerometer
▪ However for some brand new research on new applications we have
found it more useful than accelerometer
7/25/2016
Gary M. Weiss
DMIN Invited Talk
11

Barometer (Pressure Sensor)
 Measured in millibars
 Could be used to determine altitude
▪ Use to adjust calories burned if walking up hill
 More accurate and localized weather prediction

Light Sensor
 Measured in Luxes
 Uses:
▪ automatically adjust device display brightness
▪ Virtual proximity sensor: sense “head” and turn off screen
7/25/2016
Gary M. Weiss
DMIN Invited Talk
12

Proximity Sensor and Motion Sensor
 Use to recognize gestures w/o touching phone or
turn off phone when placed against ear

Magnetic Field Sensor
 Measured in microteslas (per axis)
 Used as a compass or detecting metal objects

Humidity Sensor
▪ Track weather and building conditions
7/25/2016
Gary M. Weiss
DMIN Invited Talk
13

Location Sensors
 GPS:
 Cell Tower Triangulation & Wifi
▪ Less power required than GPS; accuracy varies
 Uses:
▪ Navigation, context awareness (location)
7/25/2016
Gary M. Weiss
DMIN Invited Talk
14

Heart Rate Sensor
 Phones can simulate with camera by changing in
color on finger tip
 Some smartwatches directly measure pulse
 Uses:
▪ Fitness applications
▪ Monitor stress levels
7/25/2016
Gary M. Weiss
DMIN Invited Talk
15

Continuous monitoring of sensors was either not
considered or viewed as secondary
 Now recognized sensors need to run continuously
▪ Apple M7 motion coprocessor introduced in 2013

Battery power still an issue
 In almost all cases power is much more of a limiting
resource than CPU or RAM
 Typical sensor mining apps might drain the phone
battery in 8-12 hours
 Interestingly, TI SensorTag says battery lasts one year
7/25/2016
Gary M. Weiss
DMIN Invited Talk
16
What can we do with the Mobile Sensors?
7/25/2016
Gary M. Weiss
DMIN Invited Talk
17
Let’s think abstractly and not about any
specific application. What do sensors tell us?
7/25/2016
Gary M. Weiss
DMIN Invited Talk
18
They provide information about the user
They provide information about the
immediate environment
7/25/2016
Gary M. Weiss
DMIN Invited Talk
19
Mostly they tell us about the user.
Mostly they tell us what the user is doing.
This is called activity recognition.
Includes even more if “what” includes “where”.
“Eating at Chipotle” or “running in Central Park”
7/25/2016
Gary M. Weiss
DMIN Invited Talk
20

Who is the user?
 Biometric identification & identifying traits

What is the user doing?
 Activity recognition

Where and When is the user?
 Location and spatial based data mining applications
 Temporal based data mining applications

Who, What, Where, When, and Why?
 Social networking & context sensitive applications

How is the user
 Internal health information (heart rate, BP, emotional state, etc.)

Sensing the environment (not user)
 Crowdsource weather, group motions (panic), traffic
7/25/2016
Gary M. Weiss
DMIN Invited Talk
21
7/25/2016
Gary M. Weiss
DMIN Invited Talk
22

Activity recognition identifies user actions
 Walking, jogging, running, jumping, washing
dishes, playing basketball, reading, partying,
studying, eating, drinking, et.c

Context may matter
 Studying is more likely in a library
 Partying occurs in a social environment
▪ CenceMe listens for conversations
7/25/2016
Gary M. Weiss
DMIN Invited Talk
23

Context-sensitive applications
 Handle phone calls differently depending on context
 Play music to suit your activity
 Fuse with other info (GPS) for better results
▪ Can confirm you are on subway vs. traveling in a car19
 Untold new & innovative apps to make phones smarter

Tracking & Health applications
 Track overall activity; detect dangerous activity (falling)

Social applications
 Link users with similar behaviors (joggers, hunters)
7/25/2016
Gary M. Weiss
DMIN Invited Talk
24


Smartphones, Smartwatches, and Combination
A single accelerometer but custom hardware
 Pedometers (limited function); FitBit8


Dedicated accelerometers placed on various body
parts2,13,14,25
Multi-sensor solutions
 eWatch19: accelometer + light sensor, multiple locs.
 Smartbuckle: accelerometer + image sensor on belt

Use Phone but not a central component
 Motionbands10 multi-sensor/location transmits data to
smart phone for storage
7/25/2016
Gary M. Weiss
DMIN Invited Talk
25
1.
2.
Collect labeled raw time series sensor data
(training data)
Prepare data for mining
 Preprocess and transform data
3.
4.
Build classifier using classification algorithms
Deploy and use classifier
7/25/2016
Gary M. Weiss
DMIN Invited Talk
26

We show personalized models perform best
but user must provide labeled training data
 Our actitracker app supported self-training

Otherwise use universal models built from a
panel of representative users
7/25/2016
Gary M. Weiss
DMIN Invited Talk
27
Sensor data is time-series data
 Common classification algorithms expect “examples”
 Typical approach: extract higher level features using a
sliding window & generate fixed length records

 Average acceleration per axis, variance, binned
distributing, speed from GPS data, etc.
 Actitracker uses a 10 second window and no overlap15
 One other study uses ~7s window with 50% overlap4

Alternative: use time series prediction methods
 Few applications actually do this
7/25/2016
Gary M. Weiss
DMIN Invited Talk
28
7/25/2016
Gary M. Weiss
DMIN Invited Talk
29
Phone Walking Data
7/25/2016
Gary M. Weiss
DMIN Invited Talk
30
7/25/2016
Gary M. Weiss
DMIN Invited Talk
31
Z axis
7/25/2016
Gary M. Weiss
DMIN Invited Talk
32
7/25/2016
Gary M. Weiss
DMIN Invited Talk
33

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

2016 study: smartphones & smartwatches
 Good results over 18 activities (17 subjects)
 Hand-based activities including eating

In progress
 Increasing test subjects to 50-100 and more thorough
evaluation of the four sensors
▪ Phone accel, phone gyro, watch accel, watch gyro, fusion
7/25/2016
Gary M. Weiss
DMIN Invited Talk
34
17 test subjects
 18 activities total
 2 mins per activity
 20Hz Sampling Frequency

General Activities (hand-oriented)
 Dribbling Basketball
 Playing Catch with Tennis Ball
 Typing
 Handwriting
 Clapping
 Brushing Teeth
 Folding Clothes
General Activities
 Walking
 Jogging
 Climbing Stairs
 Sitting
 Standing
 Kicking Soccer Ball
Eating Activities (hand-oriented)
 Eating Pasta
 Eating Soup
 Eating Sandwich
 Eating Chips
 Drinking from a Cup
Gary M. Weiss, Jessica L. Timko, Catherine M. Gallagher, Kenichi Yoneda, and Andrew J. Schreiber.Smartwatchbased Activity Recognition: A Machine Learning Approach, Proceedings of the 2016 IEEE International Conference
on Biomedical and Health Informatics (BHI 2016), Las Vegas, NV, 426-429.
7/25/2016
Gary M. Weiss
DMIN Invited Talk
35
Actual Class
72.4%
Accuracy
7/25/2016
Predicted Class
Walking Jogging Stairs Sitting Standing
Lying
Down
Walking
2209
46
789
2
4
0
Jogging
45
1656
148
1
0
0
Stairs
412
54
869
3
1
0
Sitting
10
0
47
553
30
241
Standing
8
0
57
6
448
3
Lying Down
5
1
7
301
13
131
Gary M. Weiss
DMIN Invited Talk
36
98.4%
accuracy
Predicted Class
Jogging
Stairs
Walking
3033
1
24
0
0
Lying
Down
0
Jogging
4
1788
4
0
0
0
Stairs
42
4
1292
1
0
0
Sitting
0
0
4
870
2
6
Standing
5
0
11
1
509
0
Lying Down
4
0
8
7
0
442
Actual Class
Walking
7/25/2016
Gary M. Weiss
Sitting Standing
DMIN Invited Talk
37
% of Records Correctly Classified
Personal
Universal
Straw
IB3 J48 NN IB3 J48
NN
Man
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
67.6
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
71.2
37.7
Overall
7/25/2016
98.4 96.6 98.7 72.4 74.9
Gary M. Weiss
DMIN Invited Talk
38
Algorithm
Phone
accel (%)
Watch
accel (%)
Watch gyro
(%)
RF
35.1
70.3
57.5
J48
24.1
59.3
49.6
IB3
22.5
62.0
49.3
NB
26.2
63.8
53.5
MLP
18.9
64.6
57.7
Average
25.3
64.0
53.5
7/25/2016
Gary M. Weiss
DMIN Invited Talk
39
Algorithm
Phone
accel (%)
Watch
accel (%)
Watch gyro
(%)
RF
75.5
93.3
79.0
J48
65.5
86.1
73.0
IB3
67.7
93.3
60.1
NB
77.1
92.7
80.2
MLP
77.0
94.2
70.0
Average
72.6
91.9
72.4
7/25/2016
Gary M. Weiss
DMIN Invited Talk
40
Random Forest
Impersonal (%)
Personal (%)
Watch
accel
Phone
accel
Watch
gyro
Watch
accel
Phone
accel
Watch
gyro
Walking
Jogging
Stairs
Sitting
Standing
Kicking
79.8
97.7
58.5
84.9
96.3
71.3
60.7
93.8
66.7
26.9
65.9
72.5
87.0
48.6
43.1
70.5
57.9
41.4
94.2
99.2
88.9
97.5
98.1
88.7
88.5
68.8
66.7
87.0
73.1
91.7
93.5
98.1
80.0
82.2
68.6
67.9
Dribbling
Catch
Typing
Handwriting
Clapping
Brush Teeth
Fold Clothes
89.3
66.0
80.4
85.2
76.3
84.5
80.8
26.1
26.1
76.9
12.9
40.9
19.2
8.3
86.0
68.9
60.8
63.1
67.9
66.2
37.8
98.7
93.3
99.4
100.0
96.9
97.3
95.0
84.8
78.3
72.0
75.9
77.3
96.2
79.2
96.9
94.6
88.6
80.5
95.6
89.6
73.1
Eat Pasta
Eat Soup
47.1
52.7
0.0
0.0
57.9
47.7
88.6
90.7
40.0
82.4
72.9
69.8
Eat Sandwich
29.0
7.1
31.1
68.9
63.0
44.2
Eat Chips
Drink
65.0
62.7
16.0
31.8
50.6
61.1
83.4
93.3
76.0
77.3
52.5
78.5
70.3
35.1
93.3
Gary M. Weiss 57.5DMIN Invited
Talk
75.5
79.0
Activity
7/25/2016
Overall
41





Commercial smartphones can recognize activities but not
nearly as well as smartwatches
Commercial smartwatches can recognize a wide variety of
activities with relatively good accuracy
Demonstrated ability to recognize hand-based activities,
including eating activities, using smartwatches
Personal models perform better than impersonal models
for smartwatch activity recognition
Smartwatch accelerometer is shown to notably
outperform the smartwatch gyroscope for activity
recognition
7/25/2016
Gary M. Weiss
DMIN Invited Talk
42

Can finer-grained activity recognition lead to
new applications?
 Tracking/monitoring eating and drinking has
potential
▪ Microinterventions
▪ Reminder to drink water




7/25/2016
Enhanced fitness tracking
Improved knowledge of how you spend your time
Advertising opportunities; better user profiling
New applications to be created/imagined?
Gary M. Weiss
DMIN Invited Talk
43
7/25/2016
Gary M. Weiss
DMIN Invited Talk
44

Biometrics concerns unique identification
based on physical or behavioral traits
 Hard biometrics involves traits that are sufficient
to uniquely identify a person
▪ Fingerprints, DNA, iris, etc.
 Soft biometric traits are not sufficiently
distinctive, but may help
▪ Physical traits: Sex, age, height, weight, etc.
▪ Behavioral traits: gait, clothes, travel patterns, etc.
7/25/2016
Gary M. Weiss
DMIN Invited Talk
45


Equipment getting smaller, cheaper
Biometrics needs sensors and processing
 Laptops have sensors and processing
▪ Face recognition now an option

Smart phones also have sensors & processing!
 Camera might be relevant, but so is accelerometer

Substantial work on gait based biometrics
 Much of it is vision based since can be used widely
▪ Airports, etc.
7/25/2016
Gary M. Weiss
DMIN Invited Talk
46

Numerous accelerometer-based systems that
use dedicated and/or multiple sensors
 See related work section of Cell Phone-Based
Biometric Identification16 for details

Uses for phone/watch-based systems
▪
▪
▪
▪
7/25/2016
Phone security (e.g., to automatically unlock phone)9
Automatic device customization16
To better track people for shared devices
Perhaps for secondary level of physical security
Gary M. Weiss
DMIN Invited Talk
47

WISDM 2010 Biometrics Study
 Smartphone only and 36 subjects; various activities

WISDM 2015 Biometrics Study
 Smartwatch and 59 subjects; walking only
 Added some new more sophisticated features
Used for identification and authentication
 Evaluate using 10 sec. & several min. of test data

 Longer sample classify with “Most Frequent Prediction”

Same setup as WISDM activity recognition
7/25/2016
Gary M. Weiss
DMIN Invited Talk
48
Aggregate
Walk
Jog
Up
Down
Aggregate
(Oracle)
J48
72.2
84.0
83.0
65.8
61.0
76.1
Neural Net
69.5
90.9
92.2
63.3
54.5
78.6
Straw Man
4.3
4.2
5.0
6.5
4.7
4.3
Based on 10 second test samples
Aggregate
Walk
Jog
Up
Down
Aggregate
(Oracle)
J48
36/36
36/36
31/32
31/31
28/31
36/36
Neural Net
36/36
36/36
32/32
28.5/31
25/31
36/36
Based on most frequent prediction for 5-10 minutes of data
7/25/2016
Gary M. Weiss
DMIN Invited Talk
49

Authentication results:
 Positive authentication of a user
▪ 10 second sample: ~85%
▪ Most frequent class over 5-10 min: 100%
 Negative Authentication of a user (an imposter)
▪ 10 second sample: ~96%
▪ Most frequent class over 5-10 min: 100%

Near perfect results extend for unpublished
results with 200 subjects (for id and authent.)
7/25/2016
Gary M. Weiss
DMIN Invited Talk
50
Sensor
Naïve
Bayes
Random
Forest
Rotation
Forest
MLP
Avg.
Accel.
66.8%
82.9%
84.0%
83.1%
79.2%
Gyro.
52.4%
59.0%
66.4%
70.5%
62.1%
Accuracy by Algorithm and Sensor (10 sec)
Sensor
Naïve
Bayes
Random
Forest
Rotation
Forest
MLP
Avg.
Accel.
76.0%
95.3%
94.6%
95.7%
90.4%
Gyro.
66.3%
97.5%
96.4%
91.0%
87.8%
Accuracy by Algorithm and Sensor (New Features + Filtering, 10 sec)
100% Accuracy when Majority Voting Used
7/25/2016
Gary M. Weiss
DMIN Invited Talk
51
Sensor
Naïve
Bayes
Random
Forest
Rotation
Forest
MLP
Avg.
Accel.
94.9%
98.3%
98.3%
98.0%
97.2%
Gyro.
92.9%
93.8%
94.8%
94.6%
93.8%
Accuracy by Algorithm and Sensor
100% Accuracy when Majority Voting Used
7/25/2016
Gary M. Weiss
DMIN Invited Talk
52

Collecting data for many (18+) activities and
keeping data for phone & watch and accel
and gyro for both
 Will determine how identifying each activity is
 A step in the direction of continuous passive
authentication
▪ By using a two step process, where we use activity
recognition following by using a specific activity-based
biometric model
7/25/2016
Gary M. Weiss
DMIN Invited Talk
53

Data collected from ~70 people
 Accelerometer and survey data
 Survey data includes anything we could think of
that might somehow be predictable
▪
▪
▪
▪
Sex, height, weight, age, race, handedness, disability
Type of area grew up in {rural, suburban, urban}
Shoe size, footwear type, size of heels, type of clothing
# hours academic work , # hours exercise
 Too few subjects investigate all factors
▪ Many were not predictable (maybe with more data)
7/25/2016
Gary M. Weiss
DMIN Invited Talk
54
Accuracy
Male Female
71.2%
Male
31
7
Female
12
16
Accuracy Short
83.3%
Short
15
Tall
2
Tall
5
20
Accuracy
78.9%
Light
Heavy
Light
Heavy
13
2
7
17
Results for IB3 classifier. For height and weight middle categories removed.
7/25/2016
Gary M. Weiss
DMIN Invited Talk
55

When I started the smartwatch research, it was
new and exciting
 Not new and not so exciting- commercial
applications now do basic activity recognition

Smartwatches currently good area for research
 While commercial applications exist, highly
specialized activities not really tracked.
7/25/2016
Gary M. Weiss
DMIN Invited Talk
56

Next step to exploit cheap commercial sensors
 My original motivation was to use only commercial
products, so immediately usable
▪ IOT still doesn’t have immediate market comparable to
deploying a smartphone/watch app, but a reasonable
approach given future outlook
 We are now using cheap sensors like TI SensorTag
▪ Can instrument objects for $20
▪ Example: smart chair, smart shoe, smart XXX
▪ We have some good ideas for research, do you?
7/25/2016
Gary M. Weiss
DMIN Invited Talk
57

What happens when all of your appliances,
most of your clothes, and most objects you
interact with are embedded with sensors?

What happens when your home is full of
sensors?

What happens when you can correlate
patterns between you, your home, and your
family and friends?
7/25/2016
Gary M. Weiss
DMIN Invited Talk
58

Could we learn that when you eat ice cream
you are happier but your spouse feels sick
(undiagnosed lactose intolerance) and your
kids sleep 30 minutes less?
7/25/2016
Gary M. Weiss
DMIN Invited Talk
59

Gary Weiss
 Fordham University, Bronx NY 10458
 [email protected]
 http://storm.cis.fordham.edu/~gweiss/
 Papers available from my publications page

WISDM Information
 http://www.cis.fordham.edu/wisdm/
▪ Temporarily this info is out of date
7/25/2016
Gary M. Weiss
DMIN Invited Talk
60
7/25/2016
Gary M. Weiss
DMIN Invited Talk
61
1.
Agamennoni, G., Nieto, J., and Nebot, E. 2009. Mining GPS data for extracting significant places,
Proceedings of the 2009 IEEE international conference on Robotics and Automation.
2.
Bao, L. and Intille, S.. 2004. Activity recognition from user-annotated acceleration data, Lecture Notes
Computer Science, vol. 3001, pp. 1-17.
3.
Bourke, A.K., O'Brien, J.V., and Lyons, G.M. 2007. Evaluation of threshold-based tri-axial accelerometer fall
detection algorithm, Gait & Posture 26(2): 194-99.
4.
Bouten, C.V., Koekkoek, K.T., Verduin, M., Kodde, R., and Janssen, J.D. 1997. A triaxial accelerometer and
portable data processing unit for the assessment of daily physical activity, IEEE Transactions on Bio-Medical
Engineering, 44(3):136-147.
5.
Brezmes, T., Rersa, M., Gorricho, J-L, and Cotrina, J. 2010. Surveillance with Alert Management System
using Conventional Cell Phones, Proceedings of the 5th International Multi-Conference on Computing in the
Global Information Technology, 121-125.
6.
Cho, Y., Nam, Y., Choi, Y-J, and Cho, W-D. 2008,.Smart-Buckle: human activity recognition using a 3-axis
accelerometer and a wearable camera, HealthNet.
7.
Discovery channel video about a Smart phone-based biometric system for securing smart phones (based
on the research in X16). The relevant portion is about 2/3 thru the video clip which contains two segments.
Url: http://watch.discoverychannel.ca/#clip370449
7/25/2016
Gary M. Weiss
DMIN Invited Talk
62
8.
FitBit. http://www.fitbit.com
9.
Frank, J., Mannor, S., and Precup, D. 2010. Activity and gait recognition with time-delay embeddings,
Proceedings of the 24th AAAI Conference on Artificial Intelligence.
10.
Gyorbiro, N., Fabian, A., and Homanyi, G. 2008. An activity recognition system for mobile phones, Mobile
Networks and Applications, 14 (1), 82-91.
11.
Ketabdar, H., and Polzehl., T. 2009. Fall and emergency detection with mobile phones, Assets '09 Proc. of
the 11th International ACM SIGACCESS Conference on Computers and Accessibility ACM, 241-42.
12.
Khetarpaul, S., Chaujan, R., Gupta, S.K., Subramaniam, L.V., and Nambiar, U. 2011. Mining GPS data to
determine interesting locations, Proceedings of the 8th International Workshop on Information Integration on
the Web.
13.
Krishnan, N., Colbry, D., Juillard, C., and Panchanathan, S. 2008. Real time human activity recognition
using tri-Axial accelerometers, In Sensors, Signals and Information Processing Workshop.
14.
Krishnan, N., and Panchanathan, S. 2008. Analysis of low resolution accelerometer data for continuous
human activity recognition, in IEEE Int. Conf. on Acoustics, Speech and Signal Processing, pp. 3337-3340.
15.
Kwapisz, J.R., Weiss, G.M., and Moore, S.A. 2010. Activity recognition using cell phone accelerometers,
Proceedings of the Fourth International Workshop on Knowledge Discovery from Sensor Data, 10-18.
7/25/2016
Gary M. Weiss
DMIN Invited Talk
63
16.
Kwapisz, J.R.,Weiss, G.M., and Moore, S.A. 2010. Cell phone-based biometric identification, Proceedings of
the IEEE Fourth International Conference on Biometrics: Theory, Applications and Systems.
17.
Li, K.A., Sohn, T.Y., Huang, S, and Griswold, W.G. 2008. PeopleTones: A System for the detection and
notification of buddy proximity on mobile phones, Proceedings of the 6th International Conference on
Mobile Systems.
18.
Lockhart, J.W., Weiss, G.M., Xue, J.C., Gallagher, S.T., Grosner, A.B., and Pulickal, T.T. 2011. Design
considerations for the WISDM smart phone-based sensor mining architecture, In Proceedings of the Fifth
International Workshop on Knowledge Discovery from Sensor Data, San Diego, CA.
19.
Maurer, U., Smailagic, A., Siewiorek, D., and Deisher, M. 2006. Activity recognition and monitoring using
multiple sensors on different body positions, In IEEE Proceedings on the International Workshop on Wearable
and Implantable Sensor Networks, 30(5).
20.
Menn, J. February 8, 2011. Smartphone shipments surpass PCs. Retrieved from
http://www.ft.com/cms/s/2/d96e3bd8-33ca-11e0-b1ed-00144feabdc0.html#axzz1L2wKclC7
21.
Miluzzo, E., Lane, N.D., Fodor, K, Peterson, R., Lu, H., Musolesi, M., Eisenman, S.B., Zheng, X., and
Campbell, A.T. 2008. Sensing meets mobile social networks: the design, implementation and evlauation of
the CenceMe application, Proceedings of the 6th ACM on Embedded Network Sensor Systems, 337-350.
7/25/2016
Gary M. Weiss
DMIN Invited Talk
64
22.
23.
24.
25.
26.
27.
28.
29.
30.
Patterson, D., Liao, L., Fox, D, and Kautz, H. 2003. Inferring high-level behavior from low-level sensors.
Lecture Notes in Computer Science, Springer-Verlag, 73-89.
Reddy, S. Mun, M. Burke, J. Estrin, D, Hansen, M. and Srivastava, M. 2010. Using mobile phones to
determine transportation modes. ACM Transaction on Sensor Networks, 6(2).
Sposaro, F., and Tyson, G. 2009. iFall: An android application for fall monitoring and response, 31st
Annual International Conference of the IEEE Engineering in Medicine and Biology Society.
Tapia, E.M., Intille, S. et al. 2007. Real-Time recognition of physical activities and their intensities using
wireless accelerometers and a heart rate monitor, In Proc. of the 2007 11th IEEE International Symposium
on Wearable Computers.
Weiss, G.M., and Lockhart, J.W. 2011. Identifying user traits by mining smart phone accelerometer data,
Proceedings of the 5th International Workshop on Knowledge Discovery from Sensor Data.
Zhang, T., Wang, J., Liu, P., and Hou, J. 2006. Fall detection by embedding an accelerometer in cellphone
and using KFD algorithm, International Journal of Computer Science and Network Security, 6(10): 277-284.
Lockhart, J.W., Pulickal, T., and Weiss, G.M. 2012. Applications of mobile activity recognition. In
Proceedings of the ACM Ubicomp International Workshop on Situation, Activity, and Goal Awareness,
Pittsburgh, PA.
Lockhart, J.W., Weiss, G.M. The Benefits of Personalized Models for Smartphone-Based Activity
Recognition, Proceedings of the 2014 SIAM International Conference on Data Mining, Society for
Industrial and Applied Mathematics, Philadelphia, PA, 614-622.
Weiss, G.M., Lockhart, J.W., Pulickal, T.T., McHugh, P., Ronan, I.H., Timko, J.L. Actitracker: A Smartphonebased Activity Recognition System for Improving Health and Well-Being, working paper.
7/25/2016
Gary M. Weiss
DMIN Invited Talk
65