Improving the Generalization of Fisherface by Training Class Selection Using SOM2
スポンサーリンク
概要
- 論文の詳細を見る
Fisherface is a popular subspace algorithm used in face recognition,and it is commonly believed superior to another technique, Eigenface, due to itsattempt to maximize the separability of training classes. However, the deriveddiscriminant subspace of the training set may not easily extend to unseenclasses, as in the case of enrollment of new subjects. In this paper, we selectsome "representative" classes for Fisherface training using a recently proposedneural network architecture SOM2. The experiment on ORL face databaseshows the proposed method can effectively reduce the performance varianceand improve the generalization of Fisherface.
- Springer Berlin / Heidelbergの論文
- 2006-00-00
Springer Berlin / Heidelberg | 論文
- Synthesis and Photocrosslinking Reaction of N-Allylcarbamoylmethyl Cellulose Leading to Hydrogel
- A case of myelodysplastic syndrome developed blastic crisis of chronic myelogenous leukemia with acquisition of major BCR/ABL
- Usefulness of magnifying endoscopic evaluation of the terminal ileum for a patient with graft-versus-host disease after allogeneic hematopoietic stem cell transplantation
- Comparison of 99mTc-annexin A5 with 18F-FDG for the detection of atherosclerosis in ApoE−/− mice
- Hepatosplenic alpha/beta T cell lymphoma presenting with cold agglutinin disease.