Volume 11 - Issue 4
Deep Adversarial Learning on Google Home devices
- Andrea Ranieri
IMATI - National Research Council of Italy
andrea.ranieri@ge.imati.cnr.it
- Davide Caputo
DIBRIS - University of Genova, Italy
davide.caputo@dibris.unige.it
- Luca Verderame
DIBRIS - University of Genova, Italy
luca.verderame@dibris.unige.it
- Alessio Merlo
DIBRIS - University of Genova, Italy
alessio@dibris.unige.it
- Luca Caviglione
IMATI - National Research Council of Italy
luca.caviglione@ge.imati.cnr.it
Keywords: Smart Speakers, IoT privacy, Deep Adversarial Learning, Machine Learning.
Abstract
Smart speakers and voice-based virtual assistants are core components for the success of the IoT
paradigm. Unfortunately, they are vulnerable to various privacy threats exploiting machine learning
to analyze the generated encrypted traffic. To cope with that, deep adversarial learning approaches
can be used to build black-box countermeasures altering the network traffic (e.g., via packet padding)
and its statistical information. This letter showcases the inadequacy of such countermeasures against
machine learning attacks with a dedicated experimental campaign on a real network dataset. Results
indicate the need for a major re-engineering to guarantee the suitable protection of commercially
available smart speakers.