730 results about "Speech model" patented technology

The invention relates to a method and apparatus for training a multilingual speech model set. The multilingual speech model set generated is suitable for use by a speech recognition system for recognizing spoken utterances for at least two different languages. The invention allows using a single speech recognition unit with a single speech model set to perform speech recognition on utterances from two or more languages. The method and apparatus make use of a group of a group of acoustic sub-word units comprised of a first subgroup of acoustic sub-word units associated to a first language and a second subgroup of acoustic sub-word units associated to a second language where the first subgroup and the second subgroup share at least one common acoustic sub-word unit. The method and apparatus also make use of a plurality of letter to acoustic sub-word unit rules sets, each letter to acoustic sub-word unit rules set being associated to a different language. A set of untrained speech models is trained on the basis of a training set comprising speech tokens and their associated labels in combination with the group of acoustic sub-word units and the plurality of letter to acoustic sub-word unit rules sets. The invention also provides a computer readable storage medium comprising a program element for implementing the method for training a multilingual speech model set.

An apparatus for providing dynamic vocabulary prediction for setting up a speech recognition network of resource constrained portable devices may include a recognition network element. The recognition network element may be configured to determine a confidence measure for each candidate recognized word for a current word to be recognized. The recognition network element may also be configured to select a subset of candidate recognized words as selected candidate words based on the confidence measure of each one of the candidate recognized words, and determine a recognition network for a next word to be recognized, the recognition network including likely follower words for each of the selected candidate words, e.g. using language model and highly frequently used words.

Certain disclosed methods and systems perform multiple different types of message recognition using a shared language model. Message recognition of a first type is performed responsive to a first type of message input (e.g., speech), to provide text data in accordance with both the shared language model and a first model specific to the first type of message recognition (e.g., an acoustic model). Message recognition of a second type is performed responsive to a second type of message input (e.g., handwriting), to provide text data in accordance with both the shared language model and a second model specific to the second type of message recognition (e.g., a model that determines basic units of handwriting conveyed by freehand input). Accuracy of both such message recognizers can be improved by user correction of misrecognition by either one of them. Numerous other methods and systems are also disclosed.

A speech synthesis system receives symbolic input describing an utterance to be synthesized. In one embodiment, different portions of the utterance are constructed from different sources, one of which is a speech corpus recorded from a human speaker whose voice is to be modeled. The other sources may include other human speech corpora or speech produced using Rule-Based Speech Synthesis (RBSS). At least some portions of the utterance may be constructed by modifying prototype speech units to produce adapted speech units that are contextually appropriate for the utterance. The system concatenates the adapted speech units with the other speech units to produce a speech waveform. In another embodiment, a speech unit of a speech corpus recorded from a human speaker lacks transitions at one or both of its edges. A transition is synthesized using RBSS and concatenated with the speech unit in producing a speech waveform for the utterance.

A speech data mining system for use in generating a rich transcription having utility in call center management includes a speech differentiation module differentiating between speech of interacting speakers, and a speech recognition module improving automatic recognition of speech of one speaker based on interaction with another speaker employed as a reference speaker. A transcript generation module generates a rich transcript based on recognized speech of the speakers. Focused, interactive language models improve recognition of a customer on a low quality channel using context extracted from speech of a call center operator on a high quality channel with a speech model adapted to the operator. Mined speech data includes number of interaction turns, customer frustration phrases, operator polity, interruptions, and/or contexts extracted from speech recognition results, such as topics, complaints, solutions, and resolutions. Mined speech data is useful in call center and/or product or service quality management.

Speaker recognition is attempted on input speech signals concurrently with provision of input speech signals to a speech recognition system. If a speaker is recognized, a speaker dependent model which has been trained on an enrolled speaker is supplied to the speech recognition system. If not recognized, then a speaker-independent recognition model is used or, alternatively, the new speaker is enrolled. Other speaker specific information such as a special language model, grammar, vocabulary, a dictionary, a list of names, a language and speaker dependent preferences can also be provided to improve the speech recognition function or even configure or customize the speech recognition system or the response of any system such as a computer or network controlled in response thereto. A consistency check in the form of a decision tree is preferably provided to accelerate the speaker recognition process and increase the accuracy thereof. Further training of a model and/or enrollment of additional speakers may be initiated upon completion of speaker recognition and/or adaptively upon each speaker utterance.

Text-to-speech (TTS) generation is used in conjunction with large vocabulary speech recognition to say words selected by the speech recognition. The software for performing the large vocabulary speech recognition can share speech modeling data with the TTS software. TTS or recorded audio can be used to automatically say both recognized text and the names of recognized commands after their recognition. The TTS can automatically repeats text recognized by the speech recognition after each of a succession of end of utterance detections. A user can move a cursor back or forward in recognized text, and the TTS can speak one or more words at the cursor location after each such move. The speech recognition can be used to produces a choice list of possible recognition candidates and the TTS can be used to provide spoken output of one or more of the candidates on the choice list.

A method and apparatus are provided for performing speech recognition using a dynamic vocabulary. Results from a preliminary speech recognition pass can be used to update or refine a language model in order to improve the accuracy of search results and to simplify subsequent recognition passes. This iterative process greatly reduces the number of alternative hypotheses produced during each speech recognition pass, as well as the time required to process subsequent passes, making the speech recognition process faster, more efficient and more accurate. The iterative process is characterized by the use of results from one or more data set queries, where the keys used to query the data set, as well as the queries themselves, are constructed in a manner that produces more effective language models for use in subsequent attempts at decoding a given speech signal.

An arrangement is provided for an automatic speech recognition mechanism to adapt to an adverse acoustic environment. Some of the original training data, collected from an original acoustic environment, is played back in an adverse acoustic environment. The playback data is recorded in the adverse acoustic environment to generate recorded playback data. An existing speech model is then adapted with respect to the adverse acoustic environment based on the recorded playback data and/or the original training data.

A system and method for speech recognition includes generating a set of likely hypotheses in recognizing speech, rescoring the likely hypotheses by using semantic content by employing semantic structured language models, and scoring parse trees to identify a best sentence according to the sentence's parse tree by employing the semantic structured language models to clarify the recognized speech.

Voice recognition is realized by a pattern matching with a voice pattern model, and when a large number of paraphrased words are required for one facility, such as a name of a hotel or a tourist facility, the pattern matching needs to be performed with the voice pattern models of all the paraphrased words, resulting in an enormous amount of calculation. Further, it is difficult to generate all the paraphrased words, and a large amount of labor is required. A voice recognition device includes: voice recognition means for applying the voice recognition to an input voice by using a language model and an acoustic model, and outputting a predetermined number of recognition results each including a set of a recognition score and a text representation; and N-best candidate rearrangement means for: comparing the recognition result to a morpheme dictionary held in a morpheme dictionary memory; checking whether a representation of the recognition result can be expressed by any one of combinations of the morphemes of the morpheme dictionaries; correcting the recognition score when the representation can be expressed; and rearranging an order according to the corrected recognition score so as to acquire recognition results.

Linear approximation of the background noise is applied after feature extraction and prior to speaker adaptation to allow the speaker adaptation system to adapt the speech models to the enrolling user without distortion from background noise. The linear approximation is applied in the feature domain, such as in the cepstral domain. Any adaptation technique that is commutative in the feature domain may be used.

A computer system is provided including a control module 20 and data collection module 22 which generate user interfaces enabling a user to identify a vocabulary and a number of speakers from whom utterances are to be obtained. The data collection module 22 then co-ordinates the collection of utterance data for the words in the vocabulary from these speakers and stores the data in a speaker database 24. When a satisfactory set of utterances have been collected the utterances are passed to a model generation module 25 which generates a speech model using the utterances. The speech model is stored by the model generation module 25 in a model database 26. The generated model stored within the model database 26 can then be tested using a testing module 27 and other utterances stored within the speaker database 24. If the performance of the model is unsatisfactory further or different utterances can be used to generate new models for storage within the model database 26. When a speech model is determined to be satisfactory the control module 20 can invoke the output module 28 to output a copy of the model.

A language model generation and accumulation apparatus (10) that generates and accumulates language models for speech recognition is comprised of: a higher-level N-gram generation and accumulation unit (11) that generates and accumulates a higher-level N-gram language model obtained by modeling each of a plurality of texts as a string of words including a word string class having a specific linguistic property; and a lower class dependent word N-gram generation and accumulation unit (12) that generates and accumulates a lower-level N-gram language model obtained by modeling a sequence of words included in each word string class.

An automatic conversation apparatus includes a speech recognizing unit receiving a speech signal and outputting characters/character string corresponding to the speech signal as a recognition result; a speech recognition dictionary storing unit storing a language model for determining candidates corresponding to the speech signal; a conversation database storing plural pieces of topic specifying information; a sentence analyzing unit analyzing the characters/character string outputted from the speech recognizing unit; and a conversation control unit storing a conversation history and acquiring an answer sentence based on an analysis result of the sentence analyzing unit. Speech recognizing unit includes a collating unit that outputs plural candidates based on the speech recognition dictionary storing unit; and a candidate determining unit comparing the plural candidates outputted from collating unit with topic specifying information corresponding to the conversation history with reference to the conversation database and outputs one candidate based on the comparison.

A speech recognition system includes an input device to receive voice sounds, one or more processors, and one or more storage devices storing parameters and program modules including instructions which cause the one or more processors to perform operations. The operations include extracting an acoustic feature sequence from audio waveform data converted from the voice sounds, encoding the acoustic feature sequence into a hidden vector sequence using an encoder network having encoder network parameters, predicting first output label sequence probabilities by feeding the hidden vector sequence to a decoder network having decoder network parameters, predicting second output level sequence probabilities by a hybrid network using character-base language models (LMs) and word-level LMs; and searching, using a label sequence search module, for an output label sequence having a highest sequence probability by combining the first and second output label sequence probabilities provided from the decoder network and the hybrid network.

Provided are a method and apparatus for discriminating a speech signal. The apparatus for discriminating a speech signal includes: an input signal quality improver for reducing additional noise from an acoustic signal received from outside; a first start/end-point detector for receiving the acoustic signal from the input signal quality improver and detecting an end-point of a speech signal included in the acoustic signal; a voiced-speech feature extractor for extracting voiced-speech features of the input signal included in the acoustic signal received from the first start/end-point detector; a voiced-speech/unvoiced-speech discrimination model for storing a voiced-speech model parameter corresponding to a discrimination reference of the voiced-speech feature parameter extracted from the voiced-speech feature extractor; and a voiced-speech/unvoiced-speech discriminator for discriminating a voiced-speech portion using the voiced-speech features extracted by the voiced-speech feature extractor and the voiced-speech discrimination model parameter of the voiced/unvoiced-speech discrimination model.