Note: Your custom schedule will not be saved unless you create a new account or login to an existing account.
| Paper: | SS-2.1 |
| Session: | Multi-Sensory Processing for Context-Aware Computing |
| Time: | Tuesday, May 18, 13:00 - 13:20 |
| Presentation: |
Special Session Lecture |
| Topic: |
Special Sessions: Multi-sensory Processing for Context-Aware Computing |
| Title: |
OFFICE PRESENCE DETECTION USING MULTIMODAL CONTEXT INFORMATION |
| Authors: |
Xiao Huang; Michigan State University | | |
| | Juyang Weng; Michigan State University | | |
| | Zhengyou Zhang; Microsoft Research | | |
| Abstract: |
An office presence detection system is presented in this paper. Context information from multi-sensory inputs is integrated to infer a user's activities in an office. We design a layeredarchitecture to model human activities with differentgranularities. An IHDR (Incremental Hierarchical DiscriminantRegression) tree is used to automatically generate models foracoustic signals from unsegmented auditory streams, with a highadaptive capability to new settings. Hidden Markov Models (HMM)are implemented to detect human motion patterns. The outputs ofthe above two components are feed into high-level HMMs to analyzehuman activities. Experimental results of the real-time prototypesystem are reported. |
| |
| Back | |