Abstract | ||
---|---|---|
In this paper, we used hybrid features together the Bayesian network for face recognition. Hybrid features are based on Contourlet coefficients in local regions and the whole face. The combination of global features and local features achieved high accuracy. For local features, we use properties of the regions such as eyes, mouth. For global, we use the features of the whole face. The properties are converted to Contourlet domain to keep directions as well as important key points. Bayesian Network based on graph theory and probability, provide a natural tool for two problems: uncertainty and complexity and flexible, fit more datasets. The experimental results on the Olivetti Research Laboratory (ORL) [1] face dataset show that the proposed method can effectively and greatly increase the recognition. |
Year | DOI | Venue |
---|---|---|
2017 | 10.1145/3177404.3177407 | PROCEEDINGS OF 2017 INTERNATIONAL CONFERENCE ON VIDEO AND IMAGE PROCESSING (ICVIP 2017) |
Keywords | Field | DocType |
Face recognition, Bayesian network, Contourlet transform, global and local features | Graph theory,Facial recognition system,Pattern recognition,Computer science,Bayesian network,Artificial intelligence,Contourlet | Conference |
Citations | PageRank | References |
0 | 0.34 | 5 |
Authors | ||
3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Hoang Van Truong | 1 | 0 | 0.34 |
Ngo Quoc Viet | 2 | 0 | 0.34 |
Pham The Bao | 3 | 22 | 7.70 |