Electronics and Telecommunications Research Institute

Hoon Chung

Jeon Gue Park

Byung Ok Kang

Sung Joo Lee

Jeom Ja Kang

Yun Keun Lee

Ho-Young Jung

Yunkeun Lee

Hyung-Bae Jeon

Ji Hyun Wang

Jong Jin Kim

Hyung Bae Jeon

Ki-young Park

Euisok Chung

Ho Young Jung

Kap Kee Kim

Eui Sok Chung

Hwa Jeon Song

Ki Young Park

Yun-Keun Lee

Yoonhyung Kim

Sang Kyu Park

Jun Soo Park

Young Jik Lee

Seung Hi Kim

Eui-Sok Chung

Jeon-Gue Park

Sung-Joo Lee

Hoon Young Cho

Hoon-Young Cho

Provided is a self-supervised learning method based on permutation invariant cross entropy. A self-supervised learning method based on permutation invariant cross entropy performed by an electronic device includes: defining a cross entropy loss function for pre-training of an end-to-end speech recognition model; configuring non-transcription speech corpus data composed only of speech as input data of the cross entropy loss function; setting all permutations of classes included in the non-transcription speech corpus data as an output target and calculating cross entropy losses for each class; and determining a minimum cross entropy loss among the calculated cross entropy losses for each class as a final loss.

Self-supervised learning method based on permutation invariant cross entropy and electronic device thereof

Disclosed herein are an apparatus and method for audio-video sampling frequency ratio unification, including memory configured to store at least one program, and a processor configured to execute the program, wherein the program is configured to perform receiving an audio signal and a video signal, adjusting a ratio of a sampling frequency of the audio signal to a sampling frequency of the video signal so that the sampling frequency ratio is constant based on a deep learning network, and outputting an adjusted audio signal and the video signal.

Apparatus and method for audio-video sampling frequency ratio unification

Provided are an apparatus and method for linearly approximating a deep neural network (DNN) model which is a non-linear function. In general, a DNN model shows good performance in generation or classification tasks. However, the DNN fundamentally has non-linear characteristics, and therefore it is difficult to interpret how a result from inputs given to a black box model has been derived. To solve this problem, linear approximation of a DNN is proposed. The method for linearly approximating a DNN model includes 1) converting a neuron constituting a DNN into a polynomial, and 2) classifying the obtained polynomial as a polynomial of input signals and a polynomial of weights.

Apparatus and method for linearly approximating deep neural network model

Provided are a signal processing algorithm-integrated deep neural network (DNN)-based speech recognition apparatus and a learning method thereof. A model parameter learning method in a deep neural network (DNN)-based speech recognition apparatus implementable by a computer includes converting a signal processing algorithm for extracting a feature parameter from a speech input signal of a time domain into signal processing deep neural network (DNN), fusing the signal processing DNN and a classification DNN, and learning a model parameter in a deep learning model in which the signal processing DNN and the classification DNN are fused.

Signal processing algorithm-integrated deep neural network-based speech recognition apparatus and learning method thereof

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.

Apparatus and method for large vocabulary continuous speech recognition

An apparatus and method for verifying an utterance based on multi-event detection information in a natural language speech recognition system. The apparatus includes a noise processor configured to process noise of an input speech signal, a feature extractor configured to extract features of speech data obtained through the noise processing, an event detector configured to detect events of the plurality of speech features occurring in the speech data using the noise-processed data and data of the extracted features, a decoder configured to perform speech recognition using a plurality of preset speech recognition models for the extracted feature data, and an utterance verifier configured to calculate confidence measurement values in units of words and sentences using information on the plurality of events detected by the event detector and a preset utterance verification model and perform utterance verification according to the calculated confidence measurement values.

Apparatus and method for verifying utterance in speech recognition system

Disclosed are an apparatus and a method for detecting a speech endpoint using a WFST. The apparatus in accordance with an embodiment of the present invention includes: a speech decision portion configured to receive frame units of feature vector converted from a speech signal and to analyze and classify the received feature vector into a speech class or a noise class; a frame level WFST configured to receive the speech class and the noise class and to convert the speech class and the noise class to a WFST format; a speech level WFST configured to detect a speech endpoint by analyzing a relationship between the speech class and noise class and a preset state; a WFST combination portion configured to combine the frame level WFST with the speech level WFST; and an optimization portion configured to optimize the combined WFST having the frame level WFST and the speech level WFST combined therein to have a minimum route.

Method and apparatus for detecting speech endpoint using weighted finite state transducer

Disclosed are a personalized advertisement device based on speech recognition SMS services and a personalized advertisement exposure method based on speech recognition SMS services. The present invention provides a personalized advertisement device based on speech recognition SMS services and a personalized advertisement exposure method based on speech recognition SMS services capable of maximizing an effect of advertisement by grasping user's intention, an emotion state, and positional information from speech data uttered by a user during a process of providing speech recognition SMS services, configuring advertisements from when speech data begins conversion to when it has been completely converted by the speech recognition into character strings, and exposing the configured advertisements to a user.

Personalized advertisement device based on speech recognition SMS service, and personalized advertisement exposure method based on partial speech recognition SMS service

A smart watch in accordance with an embodiment of the present invention comprises: a first smart member configured to receive a voice signal sent from a mobile terminal, transform the input voice of a user to a voice signal, and send the voice signal to the mobile terminal while in talk mode; and a second smart member configured to input a control command about the talk mode into the first smart member, and transform the voice signal to voice and output the voice.

Smart watch

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.

Method for estimating language model weight and system for the same

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.

Noise reduction for speech recognition in a moving vehicle

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.

Utterance verification method and apparatus for isolated word N-best recognition result

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.

Method for separating source signals and apparatus thereof

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. The filtered probability may be a maximum value, a mean value or a median value of the previous observation probabilities and the current observation probability.

Viterbi decoder and speech recognition method using same using non-linear filter for observation probabilities

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.

Growing community of inventors

Daejeon, South Korea

Hoon Chung