中科微点-全球专利检索

  1. Login
  2. Sign Up

Search Results

6 records (took 0.017 seconds)

    METHOD FOR CONSTITUTING SPEECH MODEL AND SPEECH RECOGNITION DEVICE
    1.
    发明专利
    METHOD FOR CONSTITUTING SPEECH MODEL AND SPEECH RECOGNITION DEVICE 失效

    公开(公告)号：JPH06110493A

    公开(公告)日：1994-04-22

    申请号：JP25930192

    申请日：1992-09-29

    Applicant: IBM JAPAN

    Inventor： NISHIMURA MASAFUMI ,  OKOCHI MASAAKI

    IPC: G10L15/06 ,  G10L15/02 ,  G10L15/14 ,  G10L15/18 ,  G10L3/00

    Abstract: PURPOSE:To provide the speech recognition device which efficiently represents various vocalization deformation with a statistical combination of a small number of kind of HMMs. CONSTITUTION:A feature extracting device 4 analyzes features of an input word to obtain a corresponding feature vector train or a label train by a labeling device 8. At every vocalization deformation candidate as a speech of a subword, a phonemic hidden Markov model is given N-gram relation (N: integer large than two) with a speech deformation candidate for a precedent subword in a word and held in a parameter table 18. The recognition device 16 applies the HMM for each speech deformation candidate on the basis of the N-gram relation corresponding to a description candidate word in a recognition object word vocalization dictionary 13, constitutes a speech model by connecting respective HMMs of respective speech deformation candidates in parallel between the subwords, and finds the probability that the constituted speech model outputs the label train or feature vector train of the speech-inputted word as to each candidate word, thereby outputting the candidate word corresponding to the speech model with the highest probability as a recognition result to a display device 19.

    METHOD AND SYSTEM FOR SPEECH SYNTHESIS
    3.
    发明专利
    METHOD AND SYSTEM FOR SPEECH SYNTHESIS 失效

    公开(公告)号：JPH08123455A

    公开(公告)日：1996-05-17

    申请号：JP25319094

    申请日：1994-10-19

    Applicant: IBM JAPAN

    Inventor： SAITO TAKASHI ,  OKOCHI MASAAKI

    IPC: G06F17/27 ,  G10L13/06 ,  G10L13/08 ,  G10L3/00 ,  G10L5/04

    Abstract: PURPOSE: To synthesize a clear and highly natural speech by the use of the Japanese speech synthesis system by increasing the reproducibility of not only phoneme information, but also phonemic information. CONSTITUTION: As for Japanese, an independent word speech and an adjunct word speech are greatly different in speech feature and the difference in speech feature between the both is clearly observed as to, specially, phonemic elements (vocalizing intensity, speed, and pitch), so a new rule synthesizing method is provided which regards an adjunct word chain unit consisting of a chain of >=1 adjunct word as a speech synthesis unit and can synthesize a speech of high naturalness. The part other than the adjunct words, i.e., independent words are constituted preferably in CV/VC units.

    DEVICE AND METHOD FOR KANA-KANJI CONVERSION
    4.
    发明专利
    DEVICE AND METHOD FOR KANA-KANJI CONVERSION 失效

    公开(公告)号：JPH02105254A

    公开(公告)日：1990-04-17

    申请号：JP25701088

    申请日：1988-10-14

    Applicant: IBM JAPAN

    Inventor： KOTSURU YASUHIRO ,  NOZAKI HIROSHI ,  OKOCHI MASAAKI

    IPC: G06F17/22

    Abstract: PURPOSE:To remarkably shorten a time with respect to the input of Japanese language by using two or more KANA(Japanese syllabary)-KANJI (Chinese character) conversion parts, and performing the re-conversion of a conversion result including an error at a clause boundary by using a conversion system different from the one used in initial conversion. CONSTITUTION:A first conversion processing part 2 performs continued clause KANA-KANJI conversion on input from a keyboard 1. At this time, the retrieval of a dictionary 5 is performed along an inputted character string, and one or plural realizable clase strings are obtained first, and after that, evaluation using a score attached on the chain of words at every continued clause conversion candidate or word. A second conversion processing part 3 performs the continued clause KANA-KANJI conversion so that the number of clauses of a KANA and KANJI mixed text obtained at a part to which a re-conversion instruction is issued can be minimized. A conversion system switching part 6 judges whether or not the re-conversion instruction is the selection of a homonym, and supplies the content of a character string buffer 4 to the processing part 3 when it is not the selection of the homonym, and delivers control to the processing part 3.

Patent Agency Ranking