Projects

« Back

Biometrics Driven Smart Environments

The goal of this research is to develop smart indoor environments that can identify and track their occupants as unobtrusively as possible and answer queries about their whereabouts. The significance of this research lies in its applicability to a number of domains including assisted living for the elderly, nursing homes for the disabled, office workplaces, department stores and eventually to larger arenas such as shopping complexes, airports, etc.

The driving question underlying this research is: how do we identify, track and answer spatio-temporal queries about the occupants in ‘smart indoor environments’ monitored by unobtrusive biometric capture devices such as cameras, microphones, etc., that are distributed across various zones of the environment? As biometric recognition (e.g. face, voice, gait recognition) is inherently inexact, we propose to develop an experimental framework that will enable us to understand and predict the effect of the choice of various biometric sensors on the overall performance (precision, recall) of a smart environment. We also propose to integrate recognition with spatio-temporal reasoning in order to enhance the overall performance of the smart environment. We have developed a unified state transition system framework that serves as a basis for integrating biometric recognition, spatio-temporal reasoning, and information-theoretic retrieval concepts to provide a robust approach to designing 'smart indoor environments'.

This research will provide an experimental framework for determining the choice of modalities of the biometric recognizers, the quality of the corresponding sensors, and the type of reasoning needed in order to achieve a specified level of overall performance of the smart environment.

Publications:

• V. Menon, B. Jayaraman, and V. Govindaraju, "Biometrics Driven Smart Environments: Abstract Framework and Evaluation", 5th International Conference on Ubiquitous Intelligence and Computing (UIC-08), Springer-Verlag, pp. 75-89, Oslo, Norway, 2008.
  
• V. Menon, B. Jayaraman, and V. Govindaraju, "Integrating Recognition and Reasoning in Smart Environments", 4th IET Conference on Intelligent Environments (IE’08), IET, Seattle, WA, 2008.
  

Most biometrics are distinctive but not necessarily unique biometric data fusion allows one-to-one mapping between data and individuals.