Although numerous devices exist to track and share exercise routines based on running and walking, these devices offer limited functionality for strength-training exercises. We introduce RecoFit, a system for automatically tracking repetitive exercises – such as weight training and calisthenics – via an arm-worn inertial sensor. Our goal is to provide real-time and post-workout feedback, with no user-specific training and no intervention during a workout. Toward this end, we address three challenges:
(1) Segmenting exercise from intermittent non-exercise periods
(2) Recognizing which exercise is being performed
(3) Counting repetitions
We present cross-validation results on our training data and results from a study assessing the final system, totaling 114 participants over 146 sessions. We achieve precision and recall greater than 95% in identifying exercise periods, recognition of 99%, 98%, and 96% on circuits of 4, 7, and 13 exercises respectively, and counting that is accurate to ±1 repetition 93% of the time. These results suggest that our approach enables a new category of fitness tracking devices.
Exercise repetition detection for resistance training based on smartphonesWookjae Maeng
Regular exercise is one of the most important factors in maintaining a good state of health. In the past, different systems have been proposed to assist people when exercising. While most of those systems focus only on cardio exercises such as running and cycling, we exploit smartphones to support leisure activities with a focus on resistance training. We describe how off-the-shelf smartphones without additional external sensors can be leveraged to capture resistance training data and to give reliable training feedback. We introduce a dynamic time warping-based algorithm to detect individual resistance training repetitions from the smartphone’s acceleration stream. We evaluate the algorithm in terms of the number of correctly recognized repetitions. Additionally, for providing feedback about the quality of repetitions, we use the duration of an individual repetition and analyze how accurately start and end times of repetitions can be detected by our algorithm. Our evaluations are based on 3,598 repetitions performed by ten volunteers exercising in two distinct scenarios, a gym and a natural environment. The results show an overall repetition miscount rate of about 1 % and overall temporal detection error of about 11 % of individual repetition duration
ICVR2013 Workshop: Designing an effective rehabilitation simulationSergi Bermudez i Badia
This is the talk I gave at the International Conference on Virtual Rehabilitation 2013, part of the following workshop: Designing an effective rehabilitation simulation (Gerard Fluet, Wendy Powell, Sergi Bermúdez i Badia, Qinyin Qiu)
After completing this workshop participants should be able to: (1) Outline the process for developing, designing and testing an effective rehabilitation simulation, (2) Describe the process of shaping human movement abilities using simulated activities, (3) Identify variables that need to be considered when designing a rehabilitation activity, (4) Evaluate an open source virtual environment or game for conversion to a rehab activity, (5) Identify the strengths and weaknesses of commercially available software platforms, (6) Describe commonly used metrics to measure simulated movement performance, (7) Describe key features of a rehab activity that can be used to evaluate changes in movement performance. Intended audience: Computer Engineers, Biomedical Engineers, Physical Therapists and Occupational Therapists with less than three years experience in the design and development of simulated rehabilitation activities.
Applied Biomechanics – a multifaceted approach to answering human movement qu...InsideScientific
Experts review the basic principles of biomechanics and how the study of human movement has evolved over time. Presenters highlight examples in applied kinematics, applied kinetics and applied neuromuscular/motor control and demonstrate how methodologies vary depending on the field of study or area of expertise.
Exercise repetition detection for resistance training based on smartphonesWookjae Maeng
Regular exercise is one of the most important factors in maintaining a good state of health. In the past, different systems have been proposed to assist people when exercising. While most of those systems focus only on cardio exercises such as running and cycling, we exploit smartphones to support leisure activities with a focus on resistance training. We describe how off-the-shelf smartphones without additional external sensors can be leveraged to capture resistance training data and to give reliable training feedback. We introduce a dynamic time warping-based algorithm to detect individual resistance training repetitions from the smartphone’s acceleration stream. We evaluate the algorithm in terms of the number of correctly recognized repetitions. Additionally, for providing feedback about the quality of repetitions, we use the duration of an individual repetition and analyze how accurately start and end times of repetitions can be detected by our algorithm. Our evaluations are based on 3,598 repetitions performed by ten volunteers exercising in two distinct scenarios, a gym and a natural environment. The results show an overall repetition miscount rate of about 1 % and overall temporal detection error of about 11 % of individual repetition duration
ICVR2013 Workshop: Designing an effective rehabilitation simulationSergi Bermudez i Badia
This is the talk I gave at the International Conference on Virtual Rehabilitation 2013, part of the following workshop: Designing an effective rehabilitation simulation (Gerard Fluet, Wendy Powell, Sergi Bermúdez i Badia, Qinyin Qiu)
After completing this workshop participants should be able to: (1) Outline the process for developing, designing and testing an effective rehabilitation simulation, (2) Describe the process of shaping human movement abilities using simulated activities, (3) Identify variables that need to be considered when designing a rehabilitation activity, (4) Evaluate an open source virtual environment or game for conversion to a rehab activity, (5) Identify the strengths and weaknesses of commercially available software platforms, (6) Describe commonly used metrics to measure simulated movement performance, (7) Describe key features of a rehab activity that can be used to evaluate changes in movement performance. Intended audience: Computer Engineers, Biomedical Engineers, Physical Therapists and Occupational Therapists with less than three years experience in the design and development of simulated rehabilitation activities.
Applied Biomechanics – a multifaceted approach to answering human movement qu...InsideScientific
Experts review the basic principles of biomechanics and how the study of human movement has evolved over time. Presenters highlight examples in applied kinematics, applied kinetics and applied neuromuscular/motor control and demonstrate how methodologies vary depending on the field of study or area of expertise.
A large scale study of daily information needs captured in situWookjae Maeng
The goal of this work is to provide a fundamental understanding of the daily information needs of people through a large-scale, in-depth, quantitative investigation. To this end, we have conducted one of the most comprehensive studies of information needs to date, spanning a 3-month period and involving more than 100 users. The study employed a contextual experience sampling method, a snippet-based diary technique using SMS technology, and an online Web diary to gather in situ insights into the types of needs that occur from day to day, how those needs are addressed, and how contextual, technological, and demographic factors impact on those needs. Our results not only complement earlier studies but also provide a new understanding of the intricacies of people’s daily information needs.
KnowMe and ShareMe: Understanding Automatically Discovered Personality Trai...Wookjae Maeng
There is much recent work on using the digital footprints left by people on social media to predict personal traits and gain a deeper understanding of individuals. Due to the veracity of social media, imperfections in prediction algorithms, and the sensitive nature of one’s personal traits, much research is still needed to better understand the effectiveness of this line of work, including users’ preferences of sharing their com- putationally derived traits. In this paper, we report a two- part study involving 256 participants, which (1) examines the feasibility and effectiveness of automatically deriving three types of personality traits from Twitter, including Big 5 per- sonality, basic human values, and fundamental needs, and (2) investigates users’ opinions of using and sharing these traits. Our findings show there is a potential feasibility of automati- cally deriving one’s personality traits from social media with various factors impacting the accuracy of models. The re- sults also indicate over 61.5% users are willing to share their derived traits in the workplace and that a number of factors significantly influence their sharing preferences. Since our findings demonstrate the feasibility of automatically infer- ring a user’s personal traits from social media, we discuss their implications for designing a new generation of privacy- preserving, hyper-personalized systems.
A large scale study of daily information needs captured in situWookjae Maeng
The goal of this work is to provide a fundamental understanding of the daily information needs of people through a large-scale, in-depth, quantitative investigation. To this end, we have conducted one of the most comprehensive studies of information needs to date, spanning a 3-month period and involving more than 100 users. The study employed a contextual experience sampling method, a snippet-based diary technique using SMS technology, and an online Web diary to gather in situ insights into the types of needs that occur from day to day, how those needs are addressed, and how contextual, technological, and demographic factors impact on those needs. Our results not only complement earlier studies but also provide a new understanding of the intricacies of people’s daily information needs.
KnowMe and ShareMe: Understanding Automatically Discovered Personality Trai...Wookjae Maeng
There is much recent work on using the digital footprints left by people on social media to predict personal traits and gain a deeper understanding of individuals. Due to the veracity of social media, imperfections in prediction algorithms, and the sensitive nature of one’s personal traits, much research is still needed to better understand the effectiveness of this line of work, including users’ preferences of sharing their com- putationally derived traits. In this paper, we report a two- part study involving 256 participants, which (1) examines the feasibility and effectiveness of automatically deriving three types of personality traits from Twitter, including Big 5 per- sonality, basic human values, and fundamental needs, and (2) investigates users’ opinions of using and sharing these traits. Our findings show there is a potential feasibility of automati- cally deriving one’s personality traits from social media with various factors impacting the accuracy of models. The re- sults also indicate over 61.5% users are willing to share their derived traits in the workplace and that a number of factors significantly influence their sharing preferences. Since our findings demonstrate the feasibility of automatically infer- ring a user’s personal traits from social media, we discuss their implications for designing a new generation of privacy- preserving, hyper-personalized systems.
Estimation of Walking rate in Complex activity recognitionEditor IJCATR
Physical activity recognition using embedded sensors has enabled by many context-aware applications in different areas. In
sequential acceleration data there is a natural dependence between observations of movement or behavior, a fact that has been largely
ignored in most analyses. In this paper, investigate the role that smart devices, including smartphones, can play in identifying activities
of daily living. Monitoring and precisely quantifying users’ physical activity with inertial measurement unit-based devices, for
instance, has also proven to be important in health management of patients affected by chronic diseases, e.g. We show that their
combination only improves the overall recognition performance when their individual performances are not very high, so that there is
room for performance improvement. We show that the system can be used accurately to monitor both feet movement and use this
result in many applications such as any playing. Time and frequency domain features of the signal were used to discriminate between
activities, it demonstrates accuracy of 93% when employing a random forest analytical approach.
HUMAN ACTIVITY TRACKING BY MOBILE PHONES THROUGH HEBBIAN LEARNINGgerogepatton
A method for human activity recognition using mobile phones is introduced. Using the accelerometer and gyroscope typically found in modern smartphones, a system that uses the proposed method is able to recognize low level activities, including athletic exercises, with high accuracy. A Hebbian learning preprocessing stage is used to render accelerometer and gyroscope signals independent to the orientation of the smartphone inside the user’s pocket. After preprocessing, a selected set of features are obtained and used for classification by a k-nearest neighbor or a multilayer perceptron. The trained algorithm achieves an accuracy of 95.3 percent when using the multilayer perceptron and tested on unknown users who are asked to perform the exercises after placing the mobile device in their pocket without any constraints on the orientation. Comparison of performance with respect to other popular methods is provided.
INSIGHTS IN EEG VERSUS HEG AND RT-FMRI NEURO FEEDBACK TRAINING FOR COGNITION ...ijaia
Innovative research technologies in the neurosciences have remarkably improved the perception of brain structure and function. The use of several neurofeedback training zechniques is broadly used for the memory and cognition augmentation as well as for several learning difficulties and AHDD rehabilitation.Author’s objective is to review cognitive enhancement techniques with the use of brain imaging intervention methods as well to evaluate the effects of these methods in the educational process. The efficiency and limitations of neurofeedback training with the use of EEG brain imaging, HEG scanning, namely NIR and PIR method and fMRI scan including rt-fMRI brain scanning technique are also
discussed. Moreover, technical and clinical details of several neurofeedback treatment approaches were also taken into consideration.
Hardware landscape from computer vision to wearable sensors, and a light intro for UX requirements to ensure adherence and engagement.
At the intersection of new sensors, big data, deep learning, gamification, behavioral medicine and human factors.
Applications benefiting from "quantitative sensorimotor training", "precision exercise", "precision physiotherapy" or whatever you are calling this, include weight and strength training, powerlifting, bodybuilding, martial arts, yoga, dance, musical instrument training, post-surgery rehabilitation for ACL tears, etc.
Alternative download link:
https://www.dropbox.com/s/wcfrzdjkn58xjdq/physio_pipeline_hw.pdf?dl=0
A 2019 update on the current role of robotics and simulation in neurosurgery with updates from the recent edition of Youman and Winn's Textbook of Neurosurgery. Videos in the presentation cannot be uploaded but can be viewed from youtube.
Deep Leg Vein Thrombosis (DVT): Meaning, Causes, Symptoms, Treatment, and Mor...The Lifesciences Magazine
Deep Leg Vein Thrombosis occurs when a blood clot forms in one or more of the deep veins in the legs. These clots can impede blood flow, leading to severe complications.
CHAPTER 1 SEMESTER V PREVENTIVE-PEDIATRICS.pdfSachin Sharma
This content provides an overview of preventive pediatrics. It defines preventive pediatrics as preventing disease and promoting children's physical, mental, and social well-being to achieve positive health. It discusses antenatal, postnatal, and social preventive pediatrics. It also covers various child health programs like immunization, breastfeeding, ICDS, and the roles of organizations like WHO, UNICEF, and nurses in preventive pediatrics.
Health Education on prevention of hypertensionRadhika kulvi
Hypertension is a chronic condition of concern due to its role in the causation of coronary heart diseases. Hypertension is a worldwide epidemic and important risk factor for coronary artery disease, stroke and renal diseases. Blood pressure is the force exerted by the blood against the walls of the blood vessels and is sufficient to maintain tissue perfusion during activity and rest. Hypertension is sustained elevation of BP. In adults, HTN exists when systolic blood pressure is equal to or greater than 140mmHg or diastolic BP is equal to or greater than 90mmHg. The
India Clinical Trials Market: Industry Size and Growth Trends [2030] Analyzed...Kumar Satyam
According to TechSci Research report, "India Clinical Trials Market- By Region, Competition, Forecast & Opportunities, 2030F," the India Clinical Trials Market was valued at USD 2.05 billion in 2024 and is projected to grow at a compound annual growth rate (CAGR) of 8.64% through 2030. The market is driven by a variety of factors, making India an attractive destination for pharmaceutical companies and researchers. India's vast and diverse patient population, cost-effective operational environment, and a large pool of skilled medical professionals contribute significantly to the market's growth. Additionally, increasing government support in streamlining regulations and the growing prevalence of lifestyle diseases further propel the clinical trials market.
Growing Prevalence of Lifestyle Diseases
The rising incidence of lifestyle diseases such as diabetes, cardiovascular diseases, and cancer is a major trend driving the clinical trials market in India. These conditions necessitate the development and testing of new treatment methods, creating a robust demand for clinical trials. The increasing burden of these diseases highlights the need for innovative therapies and underscores the importance of India as a key player in global clinical research.
Antibiotic Stewardship by Anushri Srivastava.pptxAnushriSrivastav
Stewardship is the act of taking good care of something.
Antimicrobial stewardship is a coordinated program that promotes the appropriate use of antimicrobials (including antibiotics), improves patient outcomes, reduces microbial resistance, and decreases the spread of infections caused by multidrug-resistant organisms.
WHO launched the Global Antimicrobial Resistance and Use Surveillance System (GLASS) in 2015 to fill knowledge gaps and inform strategies at all levels.
ACCORDING TO apic.org,
Antimicrobial stewardship is a coordinated program that promotes the appropriate use of antimicrobials (including antibiotics), improves patient outcomes, reduces microbial resistance, and decreases the spread of infections caused by multidrug-resistant organisms.
ACCORDING TO pewtrusts.org,
Antibiotic stewardship refers to efforts in doctors’ offices, hospitals, long term care facilities, and other health care settings to ensure that antibiotics are used only when necessary and appropriate
According to WHO,
Antimicrobial stewardship is a systematic approach to educate and support health care professionals to follow evidence-based guidelines for prescribing and administering antimicrobials
In 1996, John McGowan and Dale Gerding first applied the term antimicrobial stewardship, where they suggested a causal association between antimicrobial agent use and resistance. They also focused on the urgency of large-scale controlled trials of antimicrobial-use regulation employing sophisticated epidemiologic methods, molecular typing, and precise resistance mechanism analysis.
Antimicrobial Stewardship(AMS) refers to the optimal selection, dosing, and duration of antimicrobial treatment resulting in the best clinical outcome with minimal side effects to the patients and minimal impact on subsequent resistance.
According to the 2019 report, in the US, more than 2.8 million antibiotic-resistant infections occur each year, and more than 35000 people die. In addition to this, it also mentioned that 223,900 cases of Clostridoides difficile occurred in 2017, of which 12800 people died. The report did not include viruses or parasites
VISION
Being proactive
Supporting optimal animal and human health
Exploring ways to reduce overall use of antimicrobials
Using the drugs that prevent and treat disease by killing microscopic organisms in a responsible way
GOAL
to prevent the generation and spread of antimicrobial resistance (AMR). Doing so will preserve the effectiveness of these drugs in animals and humans for years to come.
being to preserve human and animal health and the effectiveness of antimicrobial medications.
to implement a multidisciplinary approach in assembling a stewardship team to include an infectious disease physician, a clinical pharmacist with infectious diseases training, infection preventionist, and a close collaboration with the staff in the clinical microbiology laboratory
to prevent antimicrobial overuse, misuse and abuse.
to minimize the developme
RecoFit: Using a Wearable Sensor to Find, Recognize, and Count Repetitive Exercises
1. RecoFit: Using a Wearable
Sensor to Find, Recognize,
and Count Repetitive
Exercises
+ CHI 2014
- Dan Morris et al.
/ 맹욱재
x 2016 Spring
2. UX Lab, Graduate School of Convergence Science and Technology, Seoul National Univ. 2
ABSTRACT
해결책 RecoFit: 관성 센서로 반복 운동(근력, 맨몸)을 자동으로 추적하는 팔에 차는 장치
목적 사용자가 운동 중 조작/운동 종류 선택 없이, 실시간, 운동 후 피드백 제공
방법 1) 분할(segmenting) 운동과 임시 멈춤 구간 구분
2) 인지(recognizing) 어떤 운동을 하고 있는지
3) (횟수)계산 counting 반복 운동
평가 최종 시스템 평가를 위해 114명 참가자의 146 세션 training data를 교차검증
(cross-validation)
결과
정확도
운동/비운동 구분 precision > 95% recall > 95%
운동 종류 구분 (
순환 운동 구성 수)
99% (4) 98% (7) 96% (13)
운동 횟수 측정 93%
3. UX Lab, Graduate School of Convergence Science and Technology, Seoul National Univ. 3
INTRO
규칙적인 운동의 여러 장점이 있지만[11], 꾸준히 지속하기 어렵다.[20]
-> 대부분 권장 운동량을 채우지 못한다.[5]
=>만보계로 걸음수를 자동 트래킹하면 더 걷게 된다[4]
https://www.flickr.com/photos/neoadam/15837373658
https://www.youtube.com/watch?v=4LpYyULx2T4
https://commons.wikimedia.org/wiki/
운동 데이터 전송
만보계 콘솔 액세서리 유산소 운동 기계
실내 운동 + TV걷기 & 뛰기
기회를 인식 가전제품 회사가 신체 활동 측정 기기를 출시했다.
4. UX Lab, Graduate School of Convergence Science and Technology, Seoul National Univ. 4
문제점
상용화, 보편화된 기기들이 놓친 2가지 운동 영역
1) 근력 운동(weight training) 2) 맨몸 운동(calisthenics)
일부 사용자에겐 생활 방식, 취향으로 인해 근력 운동이 유산소 운동보다 지속하기 좋음
균형 잡힌 운동을 위해 근력운동은 필수며, 미 질병관리본부의 성인 최소 권장량은 주 2
회[6]
권장량에 대한 순응도(compliance)가 유산소 운동량보다 적음[5]
INTRO
calisthenics : 기구 없이 하는 근력 운동 ex) 윗몸일으키키, 팔굽혀펴기, 팔벌려높이 뛰기
Fitbit, Fuelband가 운동중 사용되며, 전체 활동에 대한 칼로리 소모량을 제공하지만
횟수, 세트수, 시간 등의 근력 운동과 관련된 high-fidelity 정보를 주는 장치는 없음
카메라 같은 센서는 동작의 다양성이나 근력 운동의 복잡성을 처리하기 어려움
웨어러블로 근력 운동을 자동 트래킹해서
GPS watch로 달리기를 시작할 때 “set it and forget it” 된 것처럼
근력 운동 중의 조작 없이 실시간, 운동후 분석 피드백을 제공함
목적
5. UX Lab, Graduate School of Convergence Science and Technology, Seoul National Univ. 5
운동 분석
REALTED WORK
RecoFit VS What can an arm holster worn smart phone do for activity recognition?
smart phone worn on arm RecoFit
공통점
1) segmentation, recognition, counting로 나눠 접근 2)
segmentation using autocorrelation features
# of participant 7 104
thresholds for
segmentation
heuristic learned
sensor placement
variation
X
dimensionality reduction
orientation-invariant
repetition counting X false peak rejection
real-time X O
결과
segmentation 85% > 95%
recognition (# of
exercises)
94% (10) subject-
independent training
96%(13)
6. UX Lab, Graduate School of Convergence Science and Technology, Seoul National Univ. 6
myHealthAssistant
Reference, 나눔명조OTF, Regular, 15pt
운동 분석
REALTED WORK
myHealthAssistant RecoFit
sensor(position)
three accelerometers
(hand, arm, leg) inertia sensor(arm)
classifier
Bayesian classifier trained
on the mean and variance
on each accelerometer
axis
HMM, SVM
counting
combination of autocorrelation-based period estimation
and peak counting on one of the accelerometer axes
결과
segmentation X > 95%
recognition (# of
exercises)
92% (13) subject-specific
training
96%(13)
To handle non-axis- aligned movements and more complex temporal patterns (e.g. secondary peaks within
repetitions, preparatory movements) that are common in natural exercise behavior
7. UX Lab, Graduate School of Convergence Science and Technology, Seoul National Univ. 7
Significant contribution: exercise segmentation, finding exercise amidst periods of
non-exercise. gesture recognition is similar problem. “gesture spotting”
https://upload.wikimedia.org/wikipedia/commons
Gesture Spotting
REALTED WORK
Autocorrelations and Periodicity
key contributions: autocorrelation function to find regions of self-similar,
repetitive exercise. highly periodic signals
ex) tracking the pitch of a musical signal, finding abnormalities in EKG signals
8. UX Lab, Graduate School of Convergence Science and Technology, Seoul National Univ. 8
Exercise looks very similar to non-exercise
WHY IS AUTOMATIC EXERCISE ANALYSIS HARD?
The most challenging problem: actually exercising time 10% ~ 100% of workout session.
Between exercises, one walks around the workout space, socializes, stretches, rests,
drinks water, selects and retrieves equipment, etc.
Distinction obvious to a human observing, but to a wearable sensor much less clear
Magnitude alone is rarely informative
most exercises are quite slowly, strive to avoid jerky movements - high acceleration.
Amplitude of acceleration during exercise consistent with that during non- exercise
non-exercise stretches > the magnitude of most exercises
“easy cases” high-velocity exercises like jumping jacks
motion magnitude > typical non-exercise magnitude.
exception
“normal cases”
motion magnitude = typical non-exercise magnitude.
9. UX Lab, Graduate School of Convergence Science and Technology, Seoul National Univ. 9
Exercise looks very similar to non-exercise
WHY IS AUTOMATIC EXERCISE ANALYSIS HARD?
Limit of Gyroscopes
Slow pushups almost no translation of an arm-worn sensor
slow rotation of a sensor = the noise of gyroscopes.
Primary observable phenomenon - not energy on any sensed axis,
but repetitive change in the gravity axis by accelerometer,
shoulder presses, non-rotational exercise:not observed at all
This fundamental diversity characterizing exercises motivates machine learning
approach to segmentation
https://upload.wikimedia.org/wikipedia/commons/e/e2/3D_Gyroscope.png
10. UX Lab, Graduate School of Convergence Science and Technology, Seoul National Univ. 10
Exercise looks very similar to non-exercise
WHY IS AUTOMATIC EXERCISE ANALYSIS HARD?
exercise is typically more periodic(repetitive) than non-exercise
-> features; autocorrelation derive metrics of repetitiveness.
https://upload.wikimedia.org/wikipedia/commons
11. UX Lab, Graduate School of Convergence Science and Technology, Seoul National Univ. 11
Exercise looks very similar to non-exercise
WHY IS AUTOMATIC EXERCISE ANALYSIS HARD?
Another challenge: walking is the most common non-exercise activity performed during
a workout. Walking is extremely periodic,
almost impossible to heuristically describe systematic differences between walking and
exercise
-> machine learning to segmentation
https://upload.wikimedia.org/wikipedia/commons
dynamic stretching – repetitive to loosen joints or muscles, not intended as exercises per
set - is quite common during a workout. tremendous challenge to robust exercise
segmentation: separating “exercise” from “dynamic stretching” is almost a question of
semantics, but one that significantly impacts UX.
it is extremely rare for an individual to consistently perform the same dynamic stretching
movement – without changing orientation – for more than a few seconds, which
supports our use of self-similarity as a core of our feature set, and motivates the
temporal smoothing approach
12. UX Lab, Graduate School of Convergence Science and Technology, Seoul National Univ. 12
Exercise looks very similar to non-exercise
WHY IS AUTOMATIC EXERCISE ANALYSIS HARD?
robust segmentation critical to the practicality of automatic exercise analysis. At the
highest level, false positives (when the system tracks an exercise that the user did not
actually perform) or false negatives (when the system fails to “credit” the user for an
exercise) are potentially disastrous from UX perspective.
when segmentation is “correct”, the boundaries need to be precise to enable robust
performance at subsequent stages of our pipeline.
reliable counting relies heavily on accurate segmentation to ignore preparatory and post-
exercise movements, such as lying down to perform pushups, or putting weights down
after biceps curls.
https://upload.wikimedia.org/wikipedia/commons
13. UX Lab, Graduate School of Convergence Science and Technology, Seoul National Univ. 13
Variability in form
WHY IS AUTOMATIC EXERCISE ANALYSIS HARD?
Aim to require no user-specific training, variability in users’ interpretations of exercise
descriptions, and their ability to consistently execute a particular form, has a tremendous
impact on recognition accuracy
Pushups, consistent definition to most potential users, exhibits wide variation in arm
posture, pace of repetition, the temporal “shape” of the movement.
And less familiar exercises often exhibit an entirely more challenging level of variability,
where users interpret the fundamental form of the exercise differently.
End-users have some familiarity with the available exercises, not typically watch a
proscriptive video or have access to a coach who would refine their form
-> no way to address the problem of variation in form other than large-scale training data
collection with enough flexibility to elicit such variation. users in both our training data
collection and our evaluation study were given instructions
“reasonable familiarity”, but allow enough interpretation to elicit natural variation.
Instructions contained an illustrative image and a high level description for each
exercise, and experimenters did not coach or correct form during data collection
14. UX Lab, Graduate School of Convergence Science and Technology, Seoul National Univ. 14
Temporal Irregularities
WHY IS AUTOMATIC EXERCISE ANALYSIS HARD?
Simple peak-picking or zero-crossing approach will yield an accurate count
jumping jacks - high-amplitude activities
typical set of squats - “double peak” for each count,
1) autocorrelation-based period estimation
2) peak counting
highlight very challenging cases.
15. UX Lab, Graduate School of Convergence Science and Technology, Seoul National Univ. 15
Unpredictable device orientation
WHY IS AUTOMATIC EXERCISE ANALYSIS HARD?
Form factor challenge: in real world use, an arm-worn sensor naturally rotate differently
around different users’ arms, due to preference and natural fit.
“trust” the axis pointing along the arm (ex: watch always has its face pointed out, in a
readable orientation), but that the device might rotate arbitrarily around the arm.
16. UX Lab, Graduate School of Convergence Science and Technology, Seoul National Univ. 16
Hardware
TRAINING DATA COLLECTION
Right forearm, inertial sensor = 3-axis accelerometer + 3-axis gyroscope.
battery, Bluetooth radio, transmitting to PC at 50Hz
17. UX Lab, Graduate School of Convergence Science and Technology, Seoul National Univ. 17
“Natural” Environment and Procedure Design
TRAINING DATA COLLECTION
it quickly became clear that early success was the result of “robotic” behavior among
training participants: segmentation dropped to precision/recall levels close to 50% when
users were in a more natural environment
it became clear that real-world deployment required a data set that was both larger and
more natural. The practicalities of labeling activities and scaling to over 100 participants
prevented us from operating in an actual gym, so we retro-fitted a large lab space to
resemble a home gym, with appropriate décor (wallpaper, curtains, etc.), video and
audio entertainment under participants’ control, a couch for rest periods, and no
computers or experimenters visible to participants.
importance of encouraging natural variability in training data.
smaller data set (30 participants), collected in a space-constrained laboratory
environment that did not aesthetically resemble a gym, with clear instructions regarding
sequencing and form.
18. UX Lab, Graduate School of Convergence Science and Technology, Seoul National Univ. 18
Overview of Pipeline
THE RECOFIT SYSTEM
whether the user is
exercising
RecoFit’s 6-axis data at 50Hz
Which particular exercise type?
segmentation recognition
counting
How many times?
19. UX Lab, Graduate School of Convergence Science and Technology, Seoul National Univ. 19
Segmentation
THE RECOFIT SYSTEM
20. UX Lab, Graduate School of Convergence Science and Technology, Seoul National Univ. 20
Counting
THE RECOFIT SYSTEM
Counting algorithm depends on the label (e.g. “jumping jacks”), raw accelerometer data
corresponding to an exercise. Empirically, gyroscope not helpful for counting.
first compute a set of candidate peaks (local maxima).
Sort these peaks based on amplitude and loop through this sorted list,
21. UX Lab, Graduate School of Convergence Science and Technology, Seoul National Univ. 21
RESULT
22. UX Lab, Graduate School of Convergence Science and Technology, Seoul National Univ. 22
RESULT
Recognition accuracy is assessed in the
context of a circuit, and inevitably the
choice of circuit affects accuracy. A
larger number of activities or high
similarity among activities will reduce
accuracy.
23. UX Lab, Graduate School of Convergence Science and Technology, Seoul National Univ. 23
DISCUSSION AND FUTURE WORK
Intensive Strength-Training and Periodicity Breakdown
Mechanical and Form Factor Considerations
User Experience Considerations
24. UX Lab, Graduate School of Convergence Science and Technology, Seoul National Univ. 24
Implications for Generalized Activity Recognition
(1) When analyzing periodic signals, the use of independent learned models for
periodicity identification and activity recognition can increase robustness.
(2) Dimensionality reduction can increase robustness to variation in device
placement and behavioral orientation.
(3) We provide specific novel features to capture self-similarity for human motion
applications, relevant to fitness, pe- dometry, physical therapy, etc.
health benefits: improved cardiovascular fitness, reduction in the risk of obesity
entertainment and social value: improved cognitive, emotional well-being