| Abstract |
A large number of applications rely on personal sensory data from devices such as smartphones, and wearables. However, the prospect of sharing sensitive personal data often prohibits large-scale user adoption. To circumvent these issues and increase data sharing, synthetic data generation has been used as an alternative to real data sharing. The generated data should preserve only the required statistics of the real data (used by the apps to provide service) and nothing else and are used as a substitute for sensitive segments of real sensors data thus protecting privacy and resulting in improved analytics. In this paper, we take a step towards generating sensory data that are hard to distinguish from real sensory data, and make two contributions: first, we present a deep learning based architecture for synthesizing sensory data. This architecture comprises a generator model, using a stack of multiple Long-Short-Term- Memory networks and a Mixture Density Network; second, we use another LSTM network based discriminator model for distinguishing between the true and the synthesized data. Using a dataset of accelerometer traces, collected using smartphones of users doing their daily activities, we show that the deep learning based discriminator model can only distinguish between the real and synthesized traces with a maximum accuracy near to 50%. |