. Play F18 Strike Force
F18 Strike Force 12 years and 11 months ago. Play F18 Strike Force
See also
F-19 Super Hornet
References
External links
Category:Flying boats
Category:Japanese Navy aircraft squadrons
Category:Military units and formations established in 1958
Category:1958 establishments in JapanThe present invention relates to a speech recognition and, more specifically, to a method and apparatus for generating a plurality of possible candidates for a word from a speech input signal.
In general, when a spoken word is to be recognized, a predetermined time interval is set, and each of the words spoken during the predetermined time interval is recognized. For example, in the case of recognition of a word "blackmail" spoken by a speaker "S", when the speaker "S" starts speaking the word "blackmail" by a predetermined time interval, the speaker "S" is allowed to continue to speak the word "blackmail". Each time when the speaker "S" begins to speak the word "blackmail" again, the speaker "S" is allowed to continue to speak the word "blackmail" by a predetermined time interval. As a result, the speaker "S" is allowed to speak a number of words "blackmail" in succession. Thus, from the succession of words "blackmail", the word "blackmail" spoken at the starting time of the speech and the word "blackmail" spoken at the ending time of the speech is recognized.
However, in the above-mentioned conventional speech recognition system, when the word spoken by the speaker is shorter than the predetermined time interval, the word spoken by the speaker is not recognized even though the word spoken by the speaker is uttered continuously. As a result, when the word spoken by the speaker is "blackmail", the word spoken by the speaker is not recognized. Moreover, even when the word spoken by the speaker is "blackmail", when the length of the word spoken by the speaker is shorter than the predetermined time interval, the word spoken by the speaker is not recognized. As a result, in the case where the speaker is forced to speak the word with a predetermined time interval in order to detect the word, when the word is spoken continuously, the word is not recognized.
In addition, in the above-mentioned conventional speech recognition system, the system is constructed to recognize only the word previously stored in the memory. As a result, when the same word is spoken a number of
Related links:
Comments