Smart Devices are Different: Assessing and MitigatingMobile Sensing Heterogeneities for Activity Recognition

By Allan Stisen, Henrik Blunck, Sourav Bhattacharya, Thor S. Prentow, Mikkel Baun Kjærgaard, Anind K. Dey, Tobias Sonne, Mads Møller Jensen. ACM International Conference on Embedded Networked Sensor Systems, 2015

850 citations recorded by the source corpus.

Subfields: Computer Science Applications, Computer Vision and Pattern Recognition, Transportation

View Paper: https://www.semanticscholar.org/paper/021ad000c6767a794a0b1378c037a56916972e92

How researchers see this paper

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

In Computer Science Applications

In Computer Vision and Pattern Recognition

In Transportation

Citations in the corpus

A Survey of Deep Active Learning (2020, methodology): “011 Robotics [83, 96] HHAR [214], NWFD [145] Smart device [37] Foursquare, Twitter [116] Social network”
Multi-task Self-Supervised Learning for Human Activity Detection (2019, background): “The Heterogeneity Human Activity Recognition (HHAR) dataset [64] contains signals from two sensors (accelerometer and gyroscope) of smartphones and smartwatches for 6 different activities, i.”
Robust Domain Generalisation by Enforcing Distribution Invariance (2016, neutral): “In activity recognition via wearable sensors, for example, training samples may be collected under specific conditions involving device type, device placement, orientation, sampling frequency, and activity performance style [Stisen et al., 2015].”
Sim2Real Transfer for Deep Reinforcement Learning with Stochastic State Transition Delays (2020, methodology): “We include additional analysis on execution latency in Appendix B. Sampling Interval: Stisen et al. [17] demonstrate that sampling interval of accelerometers can vary widely in smartphones depending on both software and hardware characteristics.”
UTiLearn: A Personalised Ubiquitous Teaching and Learning System for Smart Societies (2017, background): “8 for the Stisen smart phone accelerometer dataset [79].”
Structural Deep Clustering Network (2020, background): “• HHAR[23]: The Heterogeneity Human Activity Recognition (HHAR) dataset contains 10299 sensor records from smart phones and smart watches.”
DeepSense: A Unified Deep Learning Framework for Time-Series Mobile Sensing Data Processing (2016, methodology): “Using these sensors, mobile devices are able to infer user activities and states [35, 39] and recognize surrounding context [31, 42].”
Using Smartphones to Collect Behavioral Data in Psychological Science: Opportunities, Practical Considerations, and Challenges (2016, neutral): “Such standardization issues even arise when a standard OS is used across devices made by different manufacturers or devices containing different makes of sensors (e.g., participants using different Android devices; Stisen et al., 2015).”
Scalable real-time classification of data streams with concept drift (2017, methodology): “17 shows the speedup of the algorithm on the HAR data set.”
Deep Learning in Human Activity Recognition with Wearable Sensors: A Review on Advances (2021, background): “During recent years, many efforts have been put into improving the generalizability of models in HAR [3, 219, 262].”
Human Activity Recognition using Inertial, Physiological and Environmental Sensors: A Comprehensive Survey (2020, methodology): “integrated into personal devices such as smartphones and smartwatches, or are integrated into clothes or other specific medical equipment [6]–[9].”
Scaling Human Activity Recognition via Deep Learning-based Domain Adaptation (2018, methodology): “Indeed, the accelerometer sensor on different smartphones (of the same model) are known to exhibit perceptible differences [3], and the AR accuracy is known to degrade [4] due to such instance-specific variations in the sensor observations across different devices.”