Unsupervised adaptation for acceleration-based activity recognition: robustness to sensor displacement and rotation

Authors:
Ricardo Chavarriaga;Hamidreza Bayati;José Del Millán
Affiliations:
Defitech Chair in Non-Invasive Brain-Computer Interface (CNBI), Center for Neuroprosthetics, School of Engineering, Ecole Polytechnique Fédérale de Lausanne, Lausanne, Switzerland 1015;Defitech Chair in Non-Invasive Brain-Computer Interface (CNBI), Center for Neuroprosthetics, School of Engineering, Ecole Polytechnique Fédérale de Lausanne, Lausanne, Switzerland 1015;Defitech Chair in Non-Invasive Brain-Computer Interface (CNBI), Center for Neuroprosthetics, School of Engineering, Ecole Polytechnique Fédérale de Lausanne, Lausanne, Switzerland 1015
Venue:
Personal and Ubiquitous Computing
Year:
2013

Citing 17
Cited 3

Principles of multivariate analysis: a user's perspective

Principles of multivariate analysis: a user's perspective
What Shall We Teach Our Pants?

ISWC '00 Proceedings of the 4th IEEE International Symposium on Wearable Computers
Realtime Online Adaptive Gesture Recognition

ICPR '00 Proceedings of the International Conference on Pattern Recognition - Volume 1
Pattern Classification (2nd Edition)

Pattern Classification (2nd Edition)
Recognition-based gesture spotting in video games

Pattern Recognition Letters
Recognizing Mimicked Autistic Self-Stimulatory Behaviors Using HMMs

ISWC '05 Proceedings of the Ninth IEEE International Symposium on Wearable Computers
Adaptive visual gesture recognition for human-robot interaction using a knowledge-based software platform

Robotics and Autonomous Systems
Covariate Shift Adaptation by Importance Weighted Cross Validation

The Journal of Machine Learning Research
Activity-Aware Computing for Healthcare

IEEE Pervasive Computing
Wearable Activity Tracking in Car Manufacturing

IEEE Pervasive Computing
Dealing with sensor displacement in motion-based onbody activity recognition systems

UbiComp '08 Proceedings of the 10th international conference on Ubiquitous computing
Anomaly detection: A survey

ACM Computing Surveys (CSUR)
Activity recognition from accelerometer data

IAAI'05 Proceedings of the 17th conference on Innovative applications of artificial intelligence - Volume 3
Unsupervised Classifier Self-Calibration through Repeated Context Occurences: Is there Robustness against Sensor Displacement to Gain?

ISWC '09 Proceedings of the 2009 International Symposium on Wearable Computers
Detecting and Rectifying Anomalies in Body Sensor Networks

BSN '11 Proceedings of the 2011 International Conference on Body Sensor Networks
Unsupervised Adaptation to On-body Sensor Displacement in Acceleration-Based Activity Recognition

ISWC '11 Proceedings of the 2011 15th Annual International Symposium on Wearable Computers
A practical approach to recognizing physical activities

PERVASIVE'06 Proceedings of the 4th international conference on Pervasive Computing

A benchmark dataset to evaluate sensor displacement in activity recognition

Proceedings of the 2012 ACM Conference on Ubiquitous Computing
On-line anomaly detection and resilience in classifier ensembles

Pattern Recognition Letters
The Opportunity challenge: A benchmark database for on-body sensor-based activity recognition

Pattern Recognition Letters

Quantified Score

Hi-index	0.00

Visualization

Abstract

A common assumption in activity recognition is that the system remains unchanged between its design and its posterior operation. However, many factors affect the data distribution between two different experimental sessions. One of these factors is the potential change in the sensor location (e.g. due to replacement or slippage) affecting the classification performance. Assuming that changes in the sensor placement mainly result in shifts in the feature distributions, we propose an unsupervised adaptive classifier that calibrates itself using an online version of expectation---maximisation. Tests using three activity recognition scenarios show that the proposed adaptive algorithm is robust against shift in the feature space due to sensor displacement and rotation. Moreover, since the method estimates the change in the feature distribution, it can also be used to roughly evaluate the reliability of the system during online operation.