ABSTRACT
The hierarchical fuzzy neural networks (HFNN) are adopted to denote the FNN designed with the strategy. This chapter discusses a potential framework of FNN based on hierarchical structures is proposed to provide the ability for robust information processing. Unsupervised and supervised strategies are integrated to train the parameters in the proposed processing framework. The chapter presents some promising application examples for biometric authentication, medical image processing, video segmentation, object recognition/detection, and multimedia content-based retrieval. HFNN have become a competitive means for many applications from low-level image processing to high-level pattern recognition. The chapter discusses application examples where HFNNs can successfully facilitate recognition tasks such as detection of bank notes, recognition of human faces, or identification of breast cancer cells. In terms of the structural design, an effective implementation of FNNs hinges upon a combination of locally distributed and hierarchical networks.
