Abstract: A method and apparatus for generating, and storing for later retrieval, selected voice messages in response to selected verbal input trigger phrases. The voice signals including predetermined trigger phrases are transmitted from an input device such as a multiline telephone to a voice recognition system. The voice recognition system associates the trigger phrases with predetermined blocks of text message and in response to a sequence of trigger phrases, creates a report. The text is then output to a voice synthesizer system for the creation of synthetic voice signals corresponding to the full text of the report and the synthetic voice signals are output to a voice storage and forward system for later retrieval. The system also provides a printed copy of the full text of the report.

Abstract: To optimize valuable processor time required for the execution of the main control program in an engine control system, execution of no more than two interruption programs in preparation for execution of a frequency calculating program is provided.

Abstract: For circumstance adaption, for example, speaker adaption, confusion coefficients between the labels of the label alphabet for initial training and those for adaption are determined by alignment of adaption speech with the corresponding initially trained Markov model. That is, each piece of adaptation speech is aligned with a corresponding initially trained Markov model by the Viterbi algorithm, and each label in the adaption speech is mapped onto one of the states of the Markov models. In respect of each adaptation lable ID, the parameter values for each initial training label of the states which are mapped onto the adaptation label in concern are accumulated and normalized to generate a confusion coefficient between each initial training label and each adaptation label. The parameter table of each Markov model is rewritten in respect of the adaptation label alphabet using the confusion coefficients.

Abstract: An augmentive communications system ("ACS") and method for enabling handicapped individuals to generate sentences and also to control external devices. The ACS may employ auditory or visual feedback to present category and word choices to the user. With a single switch the user may select categories and words from these choices which then are stored in an input buffer to permit the construction of sentences. Alternatively, the user may select choices which permit the control of external devices such as a wheelchair by manipulating the single switch. The microprocessor controlled ACS system incorporates a simplified programming language and interpreter to permit nonprogrammers to construct user programs that are tailored to the particular needs of the individual user. In addition, as the user gains proficiency in using the ACS, the user program can be easily revised to permit more efficient and sophisticated communications.

Abstract: Apparatus for detecting a fundamental frequency in speech in a changing speech environment by using adaptive statistical techniques. A statistical voice detector detects changes in the voice environment by classifiers that define certain attributes of the speech to recalculate weights that are used to combine the classifiers in making the unvoiced/voiced decision that specifies whether the speech has a fundamental frequency or not. The detector is responsive to classifiers to first calculate the average of the classifiers and then to determine the overall probability that any frame will be unvoiced. In addition, the detector forms two vectors, one vector represents the statistical average of values that an unvoiced frame's classifiers would have and the other vector represents the statistical average of the values of the classifiers for a voiced frame.

Abstract: In a speech recognition system, the prior parameters of acoustic prototype vectors are adapted to a new speaker to obtain posterior parameters by having the speaker utter a set of adaptation words. The prior parameters of an acoustic prototype vector are adapted by a weighted sum of displacement vectors obtained from the adaptation utterances. Each displacement vector is associated with one segment of an uttered adaptation word. Each displacement vector represents the distance between the associated segment of the adaptation utterance and the model corresponding to that segment. Each displacement vector is weighted by the strength of the relationship of the acoustic prototype vector to the word segment model corresponding to the displacement vector.

Abstract: A method is disclosed for determining estimates of the perceived noise masking level of audio signals as a function of frequency. By developing a randomness metric related to the euclidian distance between (i) actual frequency components amplitude and phase for each block of sampled values of the signal and (ii) predicted values for these components based on values in prior blocks, it is possible to form a tonality index which provides more detailed information useful in forming the noise masking function. Application of these techniques is illustrated in a coding and decoding context for audio recording or transmission. The noise spectrum is shaped based on a noise threshold and a tonality measure for each critical frequency-band (bark).

Abstract: A continuous speech recognition system employs a grammar tree of alternative potentially recognized word paths. A technique of tracing back through the grammar tree is utilized in determining which partial word path is common to all potential word paths. The common partial word path is deleted and words corresponding to the deleted partial word path are output as recognized words.

Abstract: An apparatus and method for correctly pronouncing proper names from text using a computer provides a dictionary which performs an initial search for the name. If the name is not in the dictionary, it is sent to a filter which either positively identifies a single language group or eliminates one or more language groups as the language group of origin for that word. When the filter cannot positively identify the language group of origin for the name, a list of possible language groups is sent to a grapheme analyzer which precedes a trigram analyzer. Using grapheme analysis, the most probable language group of origin for the name is determined and sent to a language-sensitive letter-to-sound section. In this section, the name is compared with language-sensitive rules to provide accurate phonemics and stress information for the name. The phonemics (including stress information) are sent to a voice realization unit for audio output of the name.

Abstract: A communication device may be programmed to recognize voice commands via a portable programming apparatus. An identification code is employed to access particular voice recognition information (codebook) from a repository of voice recognition codebooks container within the portable programming device. The programming device responds by transmitting the identified codebook to at least one communication device, which stores the codebook therein. Thereafter, the communication device may respond to the voice commands of that individual.

Abstract: A reference pattern renewing method includes inputting an input pattern of an utterance of a word, comparing the input pattern with reference patterns stored in a dictionary memory to thereby generate at least a first candidate having the highest degree of similarity and a second candidate having the second highest degree of similarity, and calculating a ratio of the highest degree of similarity to the second highest degree of similarity. The method further includes determining whether or not the ratio is equal to or less than a predetermined threshold value, calculating, from the input pattern, a renewed reference pattern to be substituted for a corresponding one of the reference patterns when it is determined that the ratio of the highest degree of similarity to the second highest degree of similarity is neither equal to nor less than the predetermined threshold value, and registering the renewed reference pattern in the dictionary memory instead of the corresponding one of the reference patterns.

Abstract: A voice synthesis system includes the use of a group of representative sound data for synthesizing voice data. A ROM circuit in the system includes a multilevel address system that stores starting addresses of the representative sound data. The memory capacity required for storing the representative voice data synthesized is reduced by accessing nondistinguishable data through a multilevel address system.

Abstract: A method for dividing a message into a plurality of message blocks and storing each of the plurality of message blocks on separate units within a distributed data storage system. Facilities for retrieving the information from the distributed data storage system and recreating the original message for replay are also provided. If a unit within the distributed data storage system fails, then a degredated version of the original message is recreated from the remaining units and their stored message blocks.

Abstract: A text-processing system in which the sound signal corresponding to a text is applied not only to a speech-recognition device (1) but also to a first further memory (15) for storage, start marks and end marks being allocated to the words detected by the speech-recognition device (1). These marks are related in time to the sound signals stored in the first further memory (15) and are stored in a second further memory (16). In this way a correlation is obtained between the sound signals stored in the first further memory and the individual words detected by the speech-recognition device. The detected words can be reproduced on a display device (13) thereby enabling the sound signal stored in the first further memory and corresponding to a specific detected word to be reproduced acoustically so as to verify the accuracy of a word detected by the speech-recognition device and to correct an erroneously detected word.

Abstract: A speech analysis and synthesis system operates to determine a sound source signal for the entire interval of each speech unit which is to be used for speech synthesis, according to a spectrum parameter obtained from each speech unit based on cepstrum. The sound source signal and the spectrum parameter are stored for each speech unit. Speech is synthesized according to the spectrum parameter while controlling prosody of the sound source signal. The spectrum of the synthesized speech is compensated through filtering based on cepstrum.

Abstract: Electronic speech control apparatus that includes a circuit for establishing and changing over time a subject code and an utterance index. The subject code generally indicates one of a plurality of sets of utterances. The utterance index identifies an utterance of at least one word in the set indicated by the subject code. Combined with the circuit is another circuit responsive thereto for generating the utterance identified by the utterance index in the set indicated by the subject code. The apparatus also repeatedly computes an emotion and generates an utterance representing the emotion so computed. The various utterances are made in an alternating conversational fashion in response to a person's speech. These and other aspects of apparatus and method are described in greater detail herein.

Abstract: A second approximation of the multipulse excitation signal is derived from a difference signal developed from use of a first approximation of the multipulse excitation signal. Also, spectrum parameters are weighted by a periodicity measure.

Abstract: An apparatus for audibly outputting a series of numbers represented by words and divided into blocks includes a voice data memory group for storing voice data. The voice data memory group includes a commonly used memory for storing voice data which may be utilized for any of the blocks, word ending memory for storing voice data corresponding to word endings which have a distinct intonation dependent upon for which block they are to be used and a block only memory for storing block specific voice data. A block voice selecting unit selects the memory from which the voice data is to be output from among the common use memory, word ending memory and the block only memory in accordance with the block type and the number to be output.

Abstract: Speech signals are analyzed by correlating a sequence of samples to derive a sliding average magnitude difference function (SAMDF) whereby histograms are formed which are compressed and normalized to form histogram sequences representing the speech signal for comparison and recognition.

Abstract: An information recognition system includes a recognition unit for recognizing inputted information as groups of candidate characters, a memory for storing the groups of candidate characters, a display-integrated tablet for displaying top characters in the groups of candidate characters, also a coordinate input device for indicating one of the displayed characters, and a character correcting device which replaces the top character indicated by the coordinate indicating device with another character in the group to which the indicated top character belongs are provided.