Abstract
AbstractThis paper proposes the uniform structure threshold logic network which can extract various kinds of features from the multidimensional time‐series pattern. First, the basic characteristics of the proposed network are described and discussions are made on what kinds of features the network can extract from the time‐series pattern. As a result it is seen that the network has many interesting aspects in feature extraction, e.g., flexibility in the range of patterns to be extracted and determination of the center of a pattern. Based on these results, the network is applied to the feature extraction of the character pattern and the speech spectrum pattern. The results indicate that the network can extract the line‐segment components in various directions composing the character and also the location of formant in the speech spectrum pattern. The property of the network to extract these features which are important in the recognition, will be utilized in highspeed recognition and in the LSI recognition system.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: Electronics and Communications in Japan (Part I: Communications)
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
