000185773 001__ 185773
000185773 005__ 20190812205659.0
000185773 037__ $$aCONF
000185773 245__ $$aRobust Activity Recognition Combining Anomaly Detection and Classifier Retraining
000185773 269__ $$a2013
000185773 260__ $$c2013
000185773 336__ $$aConference Papers
000185773 520__ $$aActivity recognition systems based on body-worn motion sensors suffer from a decrease in performance during the deployment and run-time phases, because of probable changes in the sensors (e.g. displacement or rotation), which is the case in many real-life scenarios (e.g. mobile phone in a pocket). Existing approaches to achieve robustness tend to sacrifice information (e.g. by rotation-invariant features) or reduce the weight of the anomalous sensors at the classifier fusion stage (adaptive fusion), ignoring data which might still be perfectly meaningful, although different from the training data. We propose to use adaptation to rebuild the classifier models of the sensors which have changed position by a two-step approach: in the first step, we run an anomaly detection algorithm to automatically detect which sensors are delivering unexpected data; subsequently, we trigger a system self-training process, so that the remaining classifiers retrain the “anomalous” sensors. We show the benefit of this approach in a real activity recognition dataset comprising data from 8 sensors to recognize locomotion. The approach achieves similar accuracy compared to the upper baseline, obtained by retraining the anomalous classifiers on the new data.
000185773 6531_ $$aAnomaly detection
000185773 6531_ $$aClassifier adaptation
000185773 6531_ $$aActivity recognition
000185773 700__ $$0242182$$g191533$$aSagha, Hesam
000185773 700__ $$aCalatroni, Alberto
000185773 700__ $$0240030$$g149175$$aMillán, José del R.
000185773 700__ $$aRoggen, Daniel
000185773 700__ $$aTröster, Gerhard
000185773 700__ $$aChavarriaga, Ricardo
000185773 7112_ $$dMay 6-9, 2013$$cMIT, Cambridge$$a10th Annual Body Sensor Networks
000185773 8564_ $$zPreprint$$yPreprint$$uhttps://infoscience.epfl.ch/record/185773/files/bsn13.pdf$$s188485
000185773 909C0 $$xU12103$$pCNBI$$0252018
000185773 909C0 $$0252517$$xU12599$$pCNP
000185773 909CO $$qGLOBAL_SET$$pconf$$pSTI$$ooai:infoscience.tind.io:185773
000185773 917Z8 $$x191533
000185773 917Z8 $$x191533
000185773 917Z8 $$x137762
000185773 937__ $$aEPFL-CONF-185773
000185773 973__ $$rREVIEWED$$sACCEPTED$$aEPFL
000185773 980__ $$aCONF