Nec Corporation

Qiongqiong Wang

Takafumi Koshinaka

Kong Aik Lee

Koji Okabe

A speech enhancement meansdetermines an enhancement mask generated based on a mask for speech enhancement, when a test utterance is input as speech data. A first hyper-parameter optimization meansdetermines, when the test utterance is input, a first hyper-parameter which is a hyper-parameter representing the degree to which the signal representing the test utterance is kept using the mask, and the first hyper-parameter which is set to take into account a downstream task that is processed using an enhanced test utterance. A mask generation meansgenerates an adaptive mask from the determined enhancement mask and the first hyper-parameter that enhances the test utterance for the downstream task. The mask generation meansgenerates the adaptive mask in which the first hyper-parameter is a power of the mask.

Hyper-parameter optimization system, method, and program

A spoofing detection apparatusincludes a multi-channel spectrogram creation unitand an evaluation unit. The multi-channel spectrogram creation unitextracts different type of spectrograms from speech data and integrates the different type of spectrograms to create a multi-channel spectrogram. The evaluation unitevaluates the created multi-channel spectrogram by applying the created multi-channel spectrogram to a classifier constructed using labeled multi-channel spectrograms as training data and classifies it to either genuine or spoof.

Neural network-based signal processing apparatus, neural network-based signal processing method, and computer-readable storage medium

Provided is a pattern recognition apparatus to provide classification robustness to any kind of domain variability. The pattern recognition apparatusbased on Neural Network (NN) includes: NN training unitthat trains an NN model to generate NN parameters, based on at least one first feature vector and at least one domain vector indicating one of subsets in a specific domain, wherein, the first feature vector is extracted from each of the subsets, the domain vector indicates an identifier corresponding to the each of the subsets; and NN verification unitthat verifies a pair of second feature vectors in the specific domain to output whether the pair indicates same individual or not, based on a target domain vector and the NN parameters.

Pattern recognition apparatus, pattern recognition method, and storage medium

Spoofing detection apparatus, spoofing detection method, and computer-readable storage medium

The speech processing apparatusincludes an air microphone speech recognition unitwhich recognizes speech from an air microphoneacquiring speech through air, a wearable microphone speech recognition unitwhich recognizes speech from a wearable microphone, a sensing unitwhich measures environmental conditions, a weight decision unitwhich calculates the weights for recognition results of the air microphone speech recognition unitand the wearable microphone speech recognition uniton the basis of the environmental conditions, and a combination unitwhich combines the recognition results outputted from the air microphone speech recognition unitand the wearable microphone speech recognition unit, using the weights.

Speech processing apparatus, method, and program

A speech feature extraction apparatusincludes a voice activity detection unitthat drops non-voice frames from frames corresponding to an input speech utterance, and calculates a posterior of being voiced for each frame, a voice activity detection process unitcalculates a function value as weights in pooling frames to produce an utterance-level feature, from a given a voice activity detection posterior, and an utterance-level feature extraction unitthat extracts an utterance-level feature, from the frame on a basis of multiple frame-level features, using the function values.

Speech feature extraction apparatus, speech feature extraction method, and computer-readable storage medium

A pattern recognition apparatus for discriminative training includes: a similarity calculator that calculates similarities among training data; a statistics calculator that calculates statistics from the similarities in accordance with current labels for the training data; and a discriminative probabilistic linear discriminant analysis (PLDA) trainer that receives the training data, the statistics of the training data, the current labels and PLDA parameters, and updates the PLDA parameters and the labels of the training data.

Pattern recognition apparatus, method, and program

A speaker recognition system includes a non-transitory computer readable medium configured to store instructions. The speaker recognition system further includes a processor connected to the non-transitory computer readable medium. The processor is configured to execute the instructions for extracting acoustic features from each frame of a plurality of frames in input speech data. The processor is configured to execute the instructions for calculating a saliency value for each frame of the plurality of frames using a first neural network (NN) based on the extracted acoustic features, wherein the first NN is a trained NN using speaker posteriors. The processor is configured to execute the instructions for extracting a speaker feature using the saliency value for each frame of the plurality of frames.

Speaker recognition system and method of using the same

The A pattern recognition apparatus using domain adaptationcomprises an estimation unit. The estimation unitestimates PLDA (Probabilistic Linear Discriminant Analysis) parameters and transformation parameters from features of a first domain data and a second domain data so as to maximize/minimize an objective function with respect to the features.

Growing community of inventors

Tokyo, Japan

Qiongqiong Wang