Abstract: Provided is an apparatus for large vocabulary continuous speech recognition (LVCSR) based on a context-dependent deep neural network hidden Markov model (CD-DNN-HMM) algorithm. The apparatus may include an extractor configured to extract acoustic model-state level information corresponding to an input speech signal from a training data model set using at least one of a first feature vector based on a gammatone filterbank signal analysis algorithm and a second feature vector based on a bottleneck algorithm, and a speech recognizer configured to provide a result of recognizing the input speech signal based on the extracted acoustic model-state level information.

Abstract: An apparatus for extracting features for speech recognition in accordance with the present invention includes: a frame forming portion configured to separate input speech signals in frame units having a prescribed size; a static feature extracting portion configured to extract a static feature vector for each frame of the speech signals; a dynamic feature extracting portion configured to extract a dynamic feature vector representing a temporal variance of the extracted static feature vector by use of a basis function or a basis vector; and a feature vector combining portion configured to combine the extracted static feature vector with the extracted dynamic feature vector to configure a feature vector stream.

Abstract: Disclosed herein is a voice-based CAPTCHA method and apparatus which can perform a CAPTCHA procedure using the voice of a human being. In the voice-based CAPTCHA) method, a plurality of uttered sounds of a user are collected. A start point and an end point of a voice from each of the collected uttered sounds are detected and then speech sections are detected. Uttered sounds of the respective detected speech sections are compared with reference uttered sounds, and then it is determined whether the uttered sounds are correctly uttered sounds. It is determined whether the uttered sounds have been made by an identical speaker if it is determined that the uttered sounds are correctly uttered sounds.

Abstract: Method of the present invention may include receiving speech feature vector converted from speech signal, performing first search by applying first language model to the received speech feature vector, and outputting word lattice and first acoustic score of the word lattice as continuous speech recognition result, outputting second acoustic score as phoneme recognition result by applying an acoustic model to the speech feature vector, comparing the first acoustic score of the continuous speech recognition result with the second acoustic score of the phoneme recognition result, outputting first language model weight when the first coustic score of the continuous speech recognition result is better than the second acoustic score of the phoneme recognition result and performing a second search by applying a second language model weight, which is the same as the output first language model, to the word lattice.

Abstract: A speech recognition system includes: a speed level classifier for measuring a moving speed of a moving object by using a noise signal at an initial time of speech recognition to determine a speed level of the moving object; a first speech enhancement unit for enhancing sound quality of an input speech signal of the speech recognition by using a Wiener filter, if the speed level of the moving object is equal to or lower than a specific level; and a second speech enhancement unit enhancing the sound quality of the input speech signal by using a Gaussian mixture model, if the speed level of the moving object is higher than the specific level. The system further includes an end point detection unit for detecting start and end points, an elimination unit for eliminating sudden noise components based on a sudden noise Gaussian mixture model.

Abstract: An utterance verification method for an isolated word N-best speech recognition result includes: calculating log likelihoods of a context-dependent phoneme and an anti-phoneme model based on an N-best speech recognition result for an input utterance; measuring a confidence score of an N-best speech-recognized word using the log likelihoods; calculating distance between phonemes for the N-best speech-recognized word; comparing the confidence score with a threshold and the distance with a predetermined mean of distances; and accepting the N-best speech-recognized word when the compared results for the confidence score and the distance correspond to acceptance.

Abstract: A method for separating a sound source from a mixed signal, includes Transforming a mixed signal to channel signals in frequency domain; and grouping several frequency bands for each channel signal to form frequency clusters. Further, the method for separating the sound source from the mixed signal includes separating the frequency clusters by applying a blind source separation to signals in frequency domain for each frequency cluster; and integrating the spectrums of the separated signal to restore the sound source in a time domain wherein each of the separated signals expresses one sound source.

Abstract: A message service method using speech recognition includes a message server recognizing a speech transmitted from a transmission terminal, generating and transmitting a recognition result of the speech and N-best results based on a confusion network to the transmission terminal; if a message is selected through the recognition result and the N-best results and an evaluation result according to accuracy of the message are decided, the transmission terminal transmitting the message and the evaluation result to a reception terminal; and the reception terminal displaying the message and the evaluation result.

Abstract: A model-based distortion compensating noise reduction apparatus for speech recognition, includes: a speech absence probability calculator for calculating the probability distribution for absence and existence of a speech using the sound absence and existence information for the frames; a noise estimation updater for estimating a more accurate noise component by updating the variance of the clean speech and noise for each frame; and a speech absence probability-based noise filter for outputting a first clean speech through the speech absence probability transmitted from the speech absence probability calculator and a first noise filter. Further, the model-based distortion compensating noise reduction apparatus includes a post probability calculator for calculating post probabilities for mixtures using a GMM containing a clean speech in the first clean speech; and a final filter designer for forming a second noise filter and outputting an improved final clean speech signal using the second noise filter.

Abstract: A Viterbi decoder includes: an observation vector sequence generator for generating an observation vector sequence by converting an input speech to a sequence of observation vectors; a local optimal state calculator for obtaining a partial state sequence having a maximum similarity up to a current observation vector as an optimal state; an observation probability calculator for obtaining, as a current observation probability, a probability for observing the current observation vector in the optimal state; a buffer for storing therein a specific number of previous observation probabilities; a non-linear filter for calculating a filtered probability by using the previous observation probabilities stored in the buffer and the current observation probability; and a maximum likelihood calculator for calculating a partial maximum likelihood by using the filtered probability.

Abstract: A noise cancellation apparatus includes a noise estimation module for receiving a noise-containing input speech, and estimating a noise therefrom to output the estimated noise; a first Wiener filter module for receiving the input speech, and applying a first Wiener filter thereto to output a first estimation of clean speech; a database for storing data of a Gaussian mixture model for modeling clean speech; and an MMSE estimation module for receiving the first estimation of clean speech and the data of the Gaussian mixture model to output a second estimation of clean speech. The apparatus further includes a final clean speech estimation module for receiving the second estimation of clean speech from the MMSE estimation module and the estimated noise from the noise estimation module, and obtaining a final Wiener filter gain therefrom to output a final estimation of clean speech by applying the final Wiener filter gain.

Abstract: A microphone-array-based speech recognition system using a blind source separation (BBS) and a target speech extraction method in the system are provided. The speech recognition system performs an independent component analysis (ICA) to separate mixed signals input through a plurality of microphone into sound-source signals, extracts one target speech spoken for speech recognition from the separated sound-source signals by using a Gaussian mixture model (GMM) or a hidden Markov Model (HMM), and automatically recognizes a desired speech from the extracted target speech. Accordingly, it is possible to obtain a high speech recognition rate even in a noise environment.

Abstract: Provided are an apparatus and method for generating a noise adaptive acoustic model including a noise adaptive discriminative adaptation method. The method includes: generating a baseline model parameter from large-capacity speech training data including various noise environments; and receiving the generated baseline model parameter and applying a discriminative adaptation method to the generated results to generate an migrated acoustic model parameter suitable for an actually applied environment.

Abstract: An apparatus for evaluating the performance of speech recognition includes a speech database for storing N-number of test speech signals for evaluation. A speech recognizer is located in an actual environment and executes the speech recognition of the test speech signals reproduced using a loud speaker from the speech database in the actual environment to produce speech recognition results. A performance evaluation module evaluates the performance of the speech recognition by comparing correct recognition results answers with the speech recognition results.

Abstract: Method of the present invention may include receiving speech feature vector converted from speech signal, performing first search by applying first language model to the received speech feature vector, and outputting word lattice and first acoustic score of the word lattice as continuous speech recognition result, outputting second acoustic score as phoneme recognition result by applying an acoustic model to the speech feature vector, comparing the first acoustic score of the continuous speech recognition result with the second acoustic score of the phoneme recognition result, outputting first language model weight when the first coustic score of the continuous speech recognition result is better than the second acoustic score of the phoneme recognition result and performing a second search by applying a second language model weight, which is the same as the output first language model, to the word lattice.

Abstract: Disclosed herein are an apparatus and method for preprocessing speech signals to perform speech recognition. The apparatus includes a voiced sound interval detection unit, a preprocessing method determination unit, and a clipping signal processing unit. The voiced sound interval detection unit detects a voiced sound interval including a voiced sound signal in a voice interval. The preprocessing method determination unit detects a clipping signal present in the voiced sound interval. The clipping signal processing unit extracts signal samples adjacent to the clipping signal, and performs interpolation on the clipping signal using the adjacent signal samples.

Abstract: A model-based distortion compensating noise reduction apparatus for speech recognition, includes: a speech absence probability calculator for calculating the probability distribution for absence and existence of a speech using the sound absence and existence information for the frames; a noise estimation updater for estimating a more accurate noise component by updating the variance of the clean speech and noise for each frame; and a speech absence probability-based noise filter for outputting a first clean speech through the speech absence probability transmitted from the speech absence probability calculator and a first noise filter. Further, the model-based distortion compensating noise reduction apparatus includes a post probability calculator for calculating post probabilities for mixtures using a GMM containing a clean speech in the first clean speech; and a final filter designer for forming a second noise filter and outputting an improved final clean speech signal using the second noise filter.

Abstract: Provided is an IPTV system using voice interface which includes a voice input device, a voice processing device, a query processing and content search device, and a content providing device. The voice processing device performs voice recognition to convert voice into a text. The voice processing device includes a voice preprocessing unit, a sound model database, a language model database, and a decoder. The voice preprocessing unit performs preprocessing which includes improving the quality of sound or removing noise for the received voice, and extracts a feature vector. The decoder converts the feature vector into a text by using a sound model and a language model. Moreover, the voice processing device stores the profile and preference of a user to provide personalized service.

Abstract: An utterance verification method for an isolated word N-best speech recognition result includes: calculating log likelihoods of a context-dependent phoneme and an anti-phoneme model based on an N-best speech recognition result for an input utterance; measuring a confidence score of an N-best speech-recognized word using the log likelihoods; calculating distance between phonemes for the N-best speech-recognized word; comparing the confidence score with a threshold and the distance with a predetermined mean of distances; and accepting the N-best speech-recognized word when the compared results for the confidence score and the distance correspond to acceptance.

Abstract: A Viterbi decoder includes: an observation vector sequence generator for generating an observation vector sequence by converting an input speech to a sequence of observation vectors; a local optimal state calculator for obtaining a partial state sequence having a maximum similarity up to a current observation vector as an optimal state; an observation probability calculator for obtaining, as a current observation probability, a probability for observing the current observation vector in the optimal state; a buffer for storing therein a specific number of previous observation probabilities; a non-linear filter for calculating a filtered probability by using the previous observation probabilities stored in the buffer and the current observation probability; and a maximum likelihood calculator for calculating a partial maximum likelihood by using the filtered probability.