Preprocessing techniques for context recognition from accelerometer data

By Davide Figo, Pedro C. Diniz, Diogo R. Ferreira, João MP Cardoso. Personal and Ubiquitous Computing, 2010

594 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Signal Processing

View Paper: https://www.semanticscholar.org/paper/be19722bde9932651101514a9f050f11bdc0ef2f

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Signal Processing

Citations in the corpus

A tutorial on human activity recognition using body-worn inertial sensors (2014, background): “Preprocessing of acceleration and gyroscope signals may involve calibration, unit conversion, normalisation, resampling, synchronisation, or signal-level fusion (see [Figo et al. 2010] for a review).”
Learning Human Identity From Motion Patterns (2015, background): “In [12] and [7], exhaustive overviews of preprocessing techniques and manual feature extraction from accelerometer data for activity recognition are given.”
Fusion of Smartphone Motion Sensors for Physical Activity Recognition (2014, unknown): “Though time domain features are computationally cheap compared to frequency domain features, due to the expensive Fourier transformation [4,8], we still chose one feature set from the frequency domain to validate our results in a generic way.”
Smart Devices are Different: Assessing and MitigatingMobile Sensing Heterogeneities for Activity Recognition (2015, methodology): “More specifically, we used all time- and frequency domain features, as listed in Table 2 and 3 of [20], respectively, with the exception of the cross-time/frequency-domain Wavelet feature.”
Deep Convolutional and LSTM Recurrent Neural Networks for Multimodal Wearable Activity Recognition (2016, background): “Most recognition systems select features from a pool of “engineered” features [12].”
User-adaptive models for recognizing food preparation activities (2013, background): “…Sebastian Stein and Stephen J. McKenna CVIP, School of Computing University of Dundee Dundee, United Kingdom {sstein,stephen}@computing.dundee.ac.uk ABSTRACT Recognizing complex activities is a challenging research problem, particularly in the presence of strong variability in the way…”
Convolutional Neural Networks for human activity recognition using mobile sensors (2014, background): “As a result, many approaches were proposed for the recognition of a wide range of activities [8], [15], [10], [23].”
Window Size Impact in Human Activity Recognition (2014, methodology): “Activities involving legs, the trunk and legs and the complete body ( i.e. , 2–8, 18, 26–33) could be part of the rehabilitation or stimulation exercises of the lower body.”
Walk detection and step counting on unconstrained smartphones (2013, background): “Wavelet transforms (CWT/DWT) [36, 54], which repeatedly correlate a ‘mother’ wavelet with the signal, temporally compressing or dilating as appropriate, are more suitable [5, 43, 57] as they can capture sudden changes in the acceleration, but are known for being computationally more expensive [12].”
DeepSense: A Unified Deep Learning Framework for Time-Series Mobile Sensing Data Processing (2016, methodology): “HHAR Baselines: •HAR-RF: This algorithm [39] selects all popular time-domain and frequency domain features from [12] and ECDF features from [15], and uses random forest as classifier.”
Multi-task Self-Supervised Learning for Human Activity Detection (2019, methodology): “Before the era of end-to-end learning, manual feature design strategies [16] such as those that employ statistical measures have been used with clustering algorithms to discover a latent group of activities [70].”