Context aware access control for home voice assistant in multi-occupant homes

Abrar S. Alrumayh, Sarah M. Lehman, Chiu C. Tan

Research output: Contribution to journalArticlepeer-review

7 Scopus citations

Abstract

Home Voice Assistants (HVAs) like the Amazon Echo or the Google Home are increasingly being used to control smart devices in a smarthome. A HVA device takes in a spoken command (“Turn off the living-room lights and television”), and converts it into commands for respective devices. This paper presents CANVAS system, an Context AwareNess for Voice ASsistants system designed for multi-occupant smarthomes. CANVAS is designed to be run on a HVA, and makes use of the sounds in the home to provide additional context information to decide whether to execute the command, prompt for confirmation, or reject the command entirely. CANVAS also includes a pictorial utility to help non-expert users configure these rules. Our experiments indicate that CANVAS has a low computational overhead, making it suitable to be executed in HVAs. Our experiments Mechanical Turk suggest that the pictorial configuration utility is also intuitive.

Original languageEnglish
Article number101196
JournalPervasive and Mobile Computing
Volume67
DOIs
StatePublished - Sep 2020

Keywords

  • Access control
  • Audio processing
  • Context-aware
  • Environmental sensing
  • Smarthome configuration management
  • Social relation sensing

Fingerprint

Dive into the research topics of 'Context aware access control for home voice assistant in multi-occupant homes'. Together they form a unique fingerprint.

Cite this