Events & News
Computer Science Colloquium
| Category | Lecture |
| Date | Monday, April 21, 2008 |
| Time | 1:00 pm |
| Location | GS 701 |
| Details | Light refreshments served in the 7th floor lobby at 12:45 PM. |
| Speaker | Fahd Al-bin-ali |
| Affiliation | Computer Science Dept, University of Arizona |
Activity-Aware Computing: Modeling of Human Activity and Behavior
An important problem in Ubiquitous Computing is detecting activities using ubiquitously deployed sensors. This dissertation demonstrates how wireless sensors can be used for real-time automatic recognition of domestic activities. The dissertation assumes that home environments in the next 20 years will support a wide range of sensing technologies that are built into smart appliances and the surrounding environment (e.g. RFID tags/readers, accelerometers, current flow sensors etc.). This dissertation also assumes that there will be an abundance of embedded CPU power in the environment that will enable fast and efficient spectral analysis and feature extraction from sensor signals. Using efficient wireless technologies such as the new Bluetooth Wibree protocol, these devices will be able to communicate their sensed data in an efficient way.
Two approaches are presented for domestic activity recognition from wireless sensors. The first approach is rule-based and logical in nature and is suitable when sensor data is not present for training. Importantly, fuzzy distributions model the uncertainty and variability in expert knowledge. The second approach is probabilistic in nature and learns by observation without human intervention. This approach uses Bayesian learning and is optimized to deal with sparse data sets. The algorithms implemented are evaluated using data sets of incremental complexity collected from real homes. During the talk, I will demonstrate how the Bayesian algorithm identifies a set of domestic activities from the sensors used in this dissertation.
The impact of this research is significant in that it may ultimately allow the development of tools that monitor the activities of individuals at fine granularities. This opens the door to a range of important applications such as identifying stereotypical behaviors in Autistic children for early interventions or combating obesity through estimating energy expenditures and promoting healthier behaviors and activities.