BROWSE
Detail
Conference Paper
Multi-modal User Recognition based on Environmental Parameters
Cited - time in 
Share 
Share
- Authors
- Hye-Jin Kim, Do Hyung Kim, Su Young Chi, Jae Yeon Lee
- Issue Date
- 2010-02
- Citation
- Korea-Japan Joint Workshop on Frontiers of Computer Vision (FCV) 2010, pp.342-345
- Language
- English
- Type
- Conference Paper
- Abstract
- Multi-modal user recognition is a key-prerequisite for security. Sensor information for a person are temporally asynchronous each other. Moreover, it is known that without definite confidence criteria, multi-modal recognition methods hardly increase recognition rate. In this paper, we present an effective fusion method of multi modal observations by adopting environmental parameters. A main characteristic of this algorithm is that a particular environmental condition can cause a low accuracy for a particular modality and the other modality does not affect such condition. Accordingly, in combining audio-visual processing results, we discriminatively choose reliable features based on environmental parameters.
- KSP Keywords
- Audio-visual processing, Confidence criteria, Environmental conditions, Environmental parameters, Fusion method, Multi-modal, Recognition Rate, Recognition method, Sensor information, User recognition