Teruhiko Ukita

A speaker-independent connected digit recognition system concatenating statistically discriminated words

A recognition system for connected digits, which uses a statistical classifier to identify words in speaker-independent continuous speech, is described. The system uses the multiple similarity method, a statistical pattern recognition technique. For evaluating word strings, the system uses a scoring method that is independent of the number of words in the strings. It is derived from the a posteriori probability that a subinterval corresponds to a correct word position, giving a word similarity value. The system evaluates a word string using dynamic programming and a parallel search procedure. Experiments for the contextual effect of the training data set, for validation of the search algorithm, and for a large quantity of unspecified speakers including 40 males and 40 females were performed. For connected digits (unknown word lengths test), the string recognition rates were 90.1%-95.1% for two, three, or four connected digits, where the equivalent word (digit) rates were 97.4%-98.4%. >

A speaker independent recognition algorithm for connected word using word boundary hypothesizer

A speaker independent recognition algorithm for connected words is described which uses a word boundary hypothesizer to reduce computational cost, as well as a robust word classifier and an effective scoring strategy. The word boundary hypothesizer predicts possible candidates for word boundaries at a variable rate which is controlled by a difference in adjacent frame spectra, obtained by bandpass filters. It reduces computational cost of the algorithm to about one-tenth, compared with a conventional approach. The word classifier uses a statistical pattern recognition technique to calculate word similarities and discriminate a word for a provisional interval between hypothesized boundaries. As the scoring strategy for evaluating possible word strings, word scores are calculated and accumulated for continuous intervals. A word score is calculated from a word similarity by an equation which models the a posteriori probability of correct word position. An experiment was performed for 35 four-connected digits uttered by ten male speakers. The string recognition rate was 93.9% (word rate = 98.4%). It was also shown that the algorithm is superior to a method which regularly skips some frames for boundary hypothesis.

A similarity value transformation method for probabilistic scoring

A method to transform a similarity measure into a probability measure which indicates the reliability of classification is shown. A statistical model for the similarity value distribution is introduced for efficient estimation from a small number of samples. It is theoretically derived that the similarity value distribution in the multiple similarity method belongs to the family of Gamma distribution under this model. Several experiments were carried out to give support to the similarity value distribution model. It is shown that the estimated posterior probability using the proposed method proves effective for pattern recognition, such as connected-digit speech recognition.<<ETX>>

Knowledge representation and reasoning for discourse understanding

Extra-linguistic knowledge is necessary for discourse understanding. In this paper, we classify the knowledge, and present a framework to describe it using frames and rules. With this framework, it is easy to represent an IS-A hierarchy, which is based on a classification by different viewpoints, and to describe functions of objects as declarative knowledge. Furthermore, to treat ambiguities in discourse understanding and to process utterances based on assumptions, the system has a world mechanism for inference. Lastly, we report the evaluation of this framework through the knowledge representation of a VCR and the conversation experiment by the dialogue system.