Abstract: Methods, systems, and computer programs are provided for training a front-end neural network (“front-end NN”) and a back-end neural network (“back-end NN”). The method includes: combining the back-end NN with the front-end NN so that an output layer of the front-end NN is also an input layer of the back-end NN to form a joint layer to thereby generate a combined NN; and training the combined NN for a speech recognition with a set of utterances as training data, a plurality of specific units in the joint layer being dropped during the training and the plurality of the specific units corresponding to one or more common frequency bands. The front-end NN may be configured to estimate clean frequency filter bank features from noisy input features; or, to estimate clean frequency filter bank features from noisy frequency filter bank input features in the same feature space.

Abstract: A method, a computer program product, and a computer system for mapping between a speech signal and a transcript of the speech signal. The computer system segments the speech signal to obtain one or more segmented speech signals and the transcript of the speech signal to obtain one or more segmented transcripts of the speech signal. The computer system generates estimated phone sequences and reference phone sequences, calculates costs of correspondences between the estimated phone sequences and the reference phone sequences, determines a series of the estimated phone sequences with a smallest cost, selects a partial series of the estimated phone sequences from the series of the estimated phone sequences, and generates mapping data which includes the partial series of the estimated phone sequences and a corresponding series of the reference phone sequences.

Abstract: A substrate includes: a wiring substrate; a lower plate disposed below the wiring substrate; a canted coil spring disposed between the lower plate and the wiring substrate; an electronic component package disposed above the wiring substrate; a sheet disposed between the wiring substrate and the electronic component package and including a plurality of connection members that connects a plurality of first electrodes provided on an upper surface of the wiring substrate and a plurality of second electrodes provided on a lower surface of the electronic component package; an upper plate disposed on the electronic component package; and a coupling member that couples the lower plate and the upper plate, wherein the lower plate, the canted coil spring, the wiring substrate, the sheet, the electronic component package, and the upper plate are laminated and fixed in this order by coupling the lower plate and the upper plate by the coupling member.

Abstract: A junction box controller controls a junction box including a circuit for vehicle travel that supplies power from a battery for vehicle travel to an electrical component for vehicle travel and a charging circuit including a positive electrode charging relay and an negative electrode charging relay that are for use in coupling and decoupling the battery for vehicle travel to and from an external charging power source in a state in which the positive electrode charging relay and the negative electrode charging relay are coupled to the circuit for vehicle travel. The junction box controller includes a temperature sensor that detects the temperature of the circuit for vehicle travel, and a controller that, when the temperature of the circuit for vehicle travel reaches or exceeds a preset threshold, controls and switches on either of the positive electrode charging relay and the negative electrode charging relay.

Abstract: A set of keywords is extracted from a query. Natural Language Processing (NLP) is performed on the query to extract a set of contextual words for a keyword from the query. For the query, a first score of a first vector is computed, where the first vector represents a first contextual word. For a first result in a result set, a first result score of a first result vector is computed, where the first vector represents a first result contextual word in a set of result contextual words corresponding to the keyword in the first result. Using the first score and the first result score, a first similarity value is computed for the first result. The first result is re-ranked relative to a second result according to the first similarity value for the first result and a second similarity value for the second result in the result set.

Abstract: A computer-implemented method for processing a speech signal, includes: identifying speech segments in an input speech signal; calculating an upper variance and a lower variance, the upper variance being a variance of upper spectra larger than a criteria among speech spectra corresponding to frames in the speech segments, the lower variance being a variance of lower spectra smaller than a criteria among the speech spectra corresponding to the frames in the speech segments; determining whether the input speech signal is a special input speech signal using a difference between the upper variance and the lower variance; and performing speech recognition of the input speech signal which has been determined to be the special input speech signal, using a special acoustic model for the special input speech signal.

Abstract: A technique for detecting a music segment in an audio signal is disclosed. A time window is set for each section in an audio signal. A maximum and a statistic of the audio signal within the time window are calculated. A density index is computed for the section using the maximum and the statistic. The density index is a measure of the statistic relative to the maximum. The section is estimated as a music segment based, at least in part, on a condition with respect to the density index.

Abstract: A computer-implemented method for training a front-end neural network (“front-end NN”) and a back-end neural network (“back-end NN”) is provided. The method includes combining the back-end neural network with the front-end neural network to form a joint layer to thereby generate a combined neural network. The method also includes training the combined neural network for a speech recognition with a set of utterances as training data, with the joint layer having a plurality of frames and each frame having a plurality of bins, and where one or more specific units in each frame are dropped during the training, each of the specific units being selected randomly or based on a bin number to which the respective unit is set within its frame, with the specific units corresponding to one or more common frequency bands.

Abstract: A computer-implemented method is provided. The computer-implemented method is performed by a speech recognition system having at least a processor. The method further includes performing a speech recognition operation on the audio signal data to decode the audio signal data into a textual representation based on the estimated sound identification information from a neural network having periodic indications and components of a frequency spectrum of the audio signal data inputted thereto. The neural network includes a plurality of fully-connected network layers having a first layer that includes a plurality of first nodes and a plurality of second nodes. The method further comprises training the neural network by initially isolating the periodic indications from the components of the frequency spectrum in the first layer by setting weights between the first nodes and a plurality of input nodes corresponding to the periodic indications to 0.

Abstract: A student neural network may be trained by a computer-implemented method, including: selecting a teacher neural network among a plurality of teacher neural networks, inputting an input data to the selected teacher neural network to obtain a soft label output generated by the selected teacher neural network, and training a student neural network with at least the input data and the soft label output from the selected teacher neural network.

Abstract: A student neural network may be trained by a computer-implemented method, including: inputting common input data to each teacher neural network among a plurality of teacher neural networks to obtain a soft label output among a plurality of soft label outputs from each teacher neural network among the plurality of teacher neural networks, and training a student neural network with the input data and the plurality of soft label outputs.

Abstract: A technique for training a neural network including an input layer, one or more hidden layers and an output layer, in which the trained neural network can be used to perform a task such as speech recognition. In the technique, a base of the neural network having at least a pre-trained hidden layer is prepared. A parameter set associated with one pre-trained hidden layer in the neural network is decomposed into a plurality of new parameter sets. The number of hidden layers in the neural network is increased by using the plurality of the new parameter sets. Pre-training for the neural network is performed.

Abstract: Embodiments are directed to a system, computer program product, and method for dynamic facet dictionary management. As one or more annotations are applied to a document collection, electronic text and associated facets are identified. Additional facets and facet values are identified and selectively applied to a knowledge base. A dictionary comprised of facets and associated facet values is constructed from the selective application. Application of the dictionary to the knowledge base identifies and returns a targeted document collection. Accordingly, facet mining and dictionary construction are dynamically applied to the knowledge base.

Abstract: A method for processing a speech signal. The method comprises obtaining a logmel feature of a speech signal. The method further includes one or more processors processing the logmel feature so that the logmel feature is normalized under a constraint that a power level of the logmel feature is kept as originally obtained. The method further includes inputting the processed logmel feature into a speech-to-text system to generate corresponding text data.

Abstract: A substrate includes: a wiring substrate; a lower plate disposed below the wiring substrate; a canted coil spring disposed between the lower plate and the wiring substrate; an electronic component package disposed above the wiring substrate; a sheet disposed between the wiring substrate and the electronic component package and including a plurality of connection members that connects a plurality of first electrodes provided on an upper surface of the wiring substrate and a plurality of second electrodes provided on a lower surface of the electronic component package; an upper plate disposed on the electronic component package; and a coupling member that couples the lower plate and the upper plate, wherein the lower plate, the canted coil spring, the wiring substrate, the sheet, the electronic component package, and the upper plate are laminated and fixed in this order by coupling the lower plate and the upper plate by the coupling member.

Abstract: A method is provided for training a neural network-based (NN-based) acoustic model. The method includes receiving, by a processor, the neural network-based (NN-based) acoustic model, trained by a one-hot scheme and having an input layer, a set of middle layers, and an original output layer. At least each of the middle layers subsequent to a first one of the middle layers have trained parameters. The method further includes stacking, by the processor, a new output layer on the original output layer of the NN-based acoustic model to form a new NN-based acoustic model. The new output layer has a same size as the original output layer. The method also includes retraining, by the processor, only the new output layer and the original output layer of the new NN-based acoustic model in the one-hot scheme, with the trained parameters of middle layers subsequent to at least the first one being fixed.

Abstract: A water electrolysis system includes a water electrolysis unit; a housing; a ventilation device configured to cause air to flow inside a space of the housing; and a heating device provided on an upstream side of a water electrolysis device in a flow path of the air. A control device controls operation of the ventilation device based on ventilation temperature information detected by a ventilation temperature sensor provided above the water electrolysis device. Furthermore, the control device controls operation of the heating device based on heating temperature information of a heating temperature sensor provided between the heating device and the water electrolysis device in the flow path.

Abstract: A technique for constructing a model supporting a plurality of domains is disclosed. In the technique, a plurality of teacher models, each of which is specialized for different one of the plurality of the domains, is prepared. A plurality of training data collections, each of which is collected for different one of the plurality of the domains, is obtained. A plurality of soft label sets is generated by inputting each training data in the plurality of the training data collections into corresponding one of the plurality of the teacher models. A student model is trained using the plurality of the soft label sets.

Abstract: A computer-implemented method is provided. The computer-implemented method is performed by a speech recognition system having at least a processor. The method includes estimating sound identification information from a neural network having periodic indications and components of a frequency spectrum of an audio signal data inputted thereto. The method further includes performing a speech recognition operation on the audio signal data to decode the audio signal data into a textual representation based on the estimated sound identification information. The neural network includes a plurality of fully-connected network layers having a first layer that includes a plurality of first nodes and a plurality of second nodes. The method further comprises training the neural network by initially isolating the periodic indications from the components of the frequency spectrum in the first layer by setting weights between the first nodes and a plurality of input nodes corresponding to the periodic indications to 0.

Abstract: A via model generation method includes: acquiring via arrangement information including a hole diameter of a via formed in a board including a plurality of wiring layers, a clearance distance between a ground conductor formed in one wiring layer of the plurality of wiring layers and the via, and a ground via distance between the via and a ground via coupled to the ground conductor; acquiring board information including a relative dielectric constant of the board; calculating a capacitance component of the via by a first electromagnetic field analysis using the hole diameter of the via, the clearance distance, and the relative dielectric constant of the board; calculating an inductance component of the via by a second electromagnetic field analysis using the hole diameter of the via, the ground via distance, and the relative dielectric constant of the board; and generating a via model including the capacitance and inductance components.