Mobile sensor data anonymization

By Mohammad Malekzadeh, Richard G. Clegg, Andrea Cavallaro, Hamed Haddadi. 2018

290 citations recorded by the source corpus.

Motion sensors such as accelerometers and gyroscopes measure the instant acceleration and rotation of a device, in three dimensions. Raw data streams from motion sensors embedded in portable and wearable devices may reveal private information about users without their awareness. For example, motion data might disclose the weight or gender of a user, or enable their re-identification. To address this problem, we propose an on-device transformation of sensor data to be shared for specific applications, such as monitoring selected daily activities, without revealing information that enables user identification. We formulate the anonymization problem using an information-theoretic approach and propose a new multi-objective loss function for training deep autoencoders. This loss function helps minimizing user-identity information as well as data distortion to preserve the application-specific utility. The training process regulates the encoder to disregard user-identifiable patterns and tunes the decoder to shape the output independently of users in the training set. The trained autoencoder can be deployed on a mobile or wearable device to anonymize sensor data even for users who are not included in the training dataset. Data from 24 users transformed by the proposed anonymizing autoencoder lead to a promising trade-off between utility and privacy, with an accuracy for activity recognition above 92% and an accuracy for user identification below 7%.

Subfields: Artificial Intelligence, Computer Science Applications, Computer Vision and Pattern Recognition

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 Science Applications

In Computer Vision and Pattern Recognition

Citations in the corpus

Privacy in Deep Learning: A Survey (2020, neutral): “isms learned from data (in a generative adversarial fashion) match the theoretically optimal ones. Information-theoretic methods are also used for the anonymization of sensory data. Malekzadeh et al. [87] propose a multi-objective loss function for training deep autoencoders to extract and obfuscate”
QUOTIENT: Two-Party Secure Neural Network Training and Prediction (2019, methodology): “For MNIST and Breast cancer datasets QUOTIENT[AMSgrad=Ours, Norm=Ours] training achieves near state of the art accuracy levels for MNIST and Breast Cancer, while we differ by at most ∼2% for German credit, MotionSense and Thyroid datasets.”
Deep Learning in Human Activity Recognition with Wearable Sensors: A Review on Advances (2021, methodology): “Similarly, the same authors have proposed another anonymizing autoencoder in [182] for classifying different activities while reducing user identification accuracy.”
Privacy and Utility Preserving Sensor-Data Transformations (2019, methodology): “Hence, we use adversarial training to approximate the mutual information by estimating the posterior distribution of the sensitive data given the released data [38].”
Deep Learning for Sensor-based Human Activity Recognition (2020, background): “raining process. For example, Iwasawa et al. [65] proposed to integrate an adversarial loss with the standard activity classification loss to minimize the user identification accuracy. The authors of [92] and [91] also adopted the similar idea to prevent privacy leakage. Their experiment results sho”