Abstract
Face localization is a previous step for face recognition and in some circumstances has proven to be a difficult task. The Heteroassociative Morphological Memories are a recently proposed neural network architecture based on the shift of the basic algebraic framework. They possess some robustness to specific noise models (erosive and dilative noise). Here we show how they can be applied to the task of face localization. © Springer-Verlag Berlin Heidelberg 2001.
Cite
CITATION STYLE
Raducanu, B., & Graña, M. (2001). On the application of heteroassociative morphological memories to face localization. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 2085 LNCS, pp. 563–570). Springer Verlag. https://doi.org/10.1007/3-540-45723-2_68
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
