Abstract
A method for constructing isomorphic context-specific connectionist networks for phoneme recognition is introduced. It is shown that such networks can be merged into a single context-modulated network that makes use of second-order unit interconnections. This is accomplished by computing a minimal basis for the set of context-specific weight vectors using the singular value decomposition algorithm. Compact networks are thus obtained in which the phoneme discrimination surfaces are modulated by phonetic context. These methods are demonstrated on a small but non-trivial vowel recognition problem. It is shown that a context-modulated network can achieve a lower error rate than a context-independent network by a factor of 7. Similar results are obtained using optimized rather than constructed networks.
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 callDisclaimer: 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.
