Abstract: A driving assistance method uses a controller to perform automated lane change to control the own vehicle in such a way that the own vehicle makes a lane change under a condition that the passenger is at least in contact with the steering wheel, the driving assistance method causing the controller to perform processing including: selecting a target lane; and when automated lane change from a current lane to the target lane includes a first automated lane change from a first lane to a second lane and a second automated lane change from the second lane to a third lane, performing lane-keeping assistance under a condition that the passenger is at least in contact with the steering wheel from an end of control of the first automated lane change until a start of control of the second automated lane change.

Abstract: Techniques for training a neural transducer-based automatic speech recognition model to be robust against background additive noise and thereby reducing insertion errors. In one aspect, a method of training an automatic speech recognition model includes: generating a modified training data set from an initial training dataset by concatenating one-word utterances with a preceding or a succeeding sentence in the initial training dataset based on a duration of silence between the one-word utterances and the preceding or the succeeding sentence; and training the automatic speech recognition model using the modified training data set.

Abstract: Methods and systems for training a neural network include clustering a full set of training data samples into specialized training clusters. Specialized teacher neural networks are trained using respective specialized training clusters of the specialized training clusters. Soft labels are generated for the full set of training data samples using the specialized teacher neural networks. A student model is trained using the full set of training data samples, the specialized training clusters, and the soft labels.

Abstract: A method for manufacturing a power storage module includes: a placing step of placing an electrode laminate on a restraint member so that a laminating direction runs along a vertical direction; a connecting step of connecting a first connection part so that a connecting portion of a second connection part with respect to a connected part is positioned on an upper side in the vertical direction than an opening; a supplying step of connecting a distal end part of a supply pipe for an electrolytic solution to the second connection part, and supplying the electrolytic solution to a space; and a conveying step of removing a recessed part of the supply pipe from the second connection part, and conveying a cell stack placed on the restraint member in a state in which the opening is connected to the first connection part.

Abstract: A driving assistance method uses a controller to perform automated lane change to control the own vehicle in such a way that the own vehicle makes a lane change under a condition that the passenger is at least in contact with the steering wheel, the driving assistance method causing the controller to perform processing including: selecting a target lane; and when automated lane change from a current lane to the target lane includes a first automated lane change from a first lane to a second lane and a second automated lane change from the second lane to a third lane, performing lane-keeping assistance under a condition that the passenger is at least in contact with the steering wheel from an end of control of the first automated lane change until a start of control of the second automated lane change.

Abstract: Systems, computer-implemented methods, and computer program products to facilitate multi-task training a recurrent neural network transducer (RNN-T) using automatic speech recognition (ASR) information are provided. According to an embodiment, a system can comprise a memory that stores computer executable components and a processor that executes the computer executable components stored in the memory. The computer executable components can include an RNN-T that can receive ASR information. The computer executable components can include a voice activity detection (VAD) model that trains the RNN-T using the ASR information, where the RNN-T can further comprise an encoder and a joint network. One or more outputs of the encoder can be integrated with the joint network and one or more outputs of the VAD model.

Abstract: The present invention provides a resin which enables the production of an organic field effect transistor element which has excellent resistance to bias stress by being used as a gate insulating film layer for an organic field effect transistor. The present invention provides a resin which comprises a repeating unit represented by Formula (1) and a repeating unit represented by Formula (2), wherein: the repeating unit represented by Formula (2) has a HOMO level of ?6.4 eV or less; and 20% by mole or more of the repeating unit represented by Formula (2) is contained relative to the total amount of the repeating unit represented by Formula (1) and the repeating unit represented by Formula (2).

Abstract: In a hydrogen production apparatus, a front end of a housing is closed by an opening/closing door. In a closed state of the opening/closing door, the upper end of the upper rod and the lower end of the lower rod in the opening/closing switching mechanism are disposed behind the first front frame piece and the second front frame piece of the housing, respectively. The axial distance between the upper end of the lower collar member and the second stopper is greater than the axial distance between the upper end of the upper collar member and the first stopper.

Abstract: A restraining jig includes: a first unit configured to apply a restraining load in the Z-axis direction to an electrode laminate via a pair of restraint members disposed at both ends in the Z-axis direction of the electrode laminate; and a second unit including a flow channel part communicating with a space by being connected to an opening, and a base part supporting the flow channel part. The second unit is disposed to be freely attached to or detached from the first unit. When the base part is fixed to at least one of the pair of restraint members by fixing parts configured to regulate movement in the X-axis direction of the second unit with respect to the first unit and fix the second unit, the flow channel part is liquid-tightly connected to the opening.

Abstract: A method, computer system, and a computer program product for audio data augmentation are provided. Sets of audio data from different sources may be obtained. A respective normalization factor for at least two sources of the different sources may be calculated. The normalization factors from the at least two sources may be mixed to determine a mixed normalization factor. A first set of the sets may be normalized by using the mixed normalization factor and to obtain training data for training an acoustic model.

Abstract: A word-level confidence score is calculated using a computerized automatic speech recognition system by computing an average of confidence levels for each character in a word and a trailing space character delineating an end of the word and the word is managed using the computerized automatic speech recognition system and using a threshold process based on the calculated word-level confidence score.

Abstract: Provided are (i) a solution for forming an organic semiconductor layer which solution has excellent coating property, (ii) an organic semiconductor which is produced with use of the solution and which has high heat resistance, (iii) a layer which contains the organic semiconductor, and (iv) an organic thin film transistor which exhibits high electrical properties. A composition containing: an organic semiconductor; and a polymer (1) having at least one unit selected from the group consisting of units represented by formulae (1-a), (1-b) and (1-c) given below. A composition containing the organic semiconductor, the polymer (1) and an organic solvent can be suitably used as a solution for forming an organic semiconductor layer.

Abstract: Techniques for training a neural transducer-based automatic speech recognition model to be robust against background additive noise and thereby reducing insertion errors. In one aspect, a method of training an automatic speech recognition model includes: generating a modified training data set from an initial training dataset by concatenating one-word utterances with a preceding or a succeeding sentence in the initial training dataset based on a duration of silence between the one-word utterances and the preceding or the succeeding sentence; and training the automatic speech recognition model using the modified training data set.

Abstract: A computer-implemented method for preparing training data for a speech recognition model is provided including obtaining a plurality of sentences from a corpus, dividing each phoneme in each sentence of the plurality of sentences into three hidden states, calculating, for each sentence of the plurality of sentences, a score based on a variation in duration of the three hidden states of each phoneme in the sentence, and sorting the plurality of sentences by using the calculated scores.

Abstract: In a method for improving generation and relevancy of search results, a processor receives a search query comprising a search term. A processor generates a document group based on the search query and at least one synonym related to the search term in a synonym dictionary. The synonym dictionary may include search document attributes for base words and synonyms of the base words. A processor extracts, from the document group, an extracted document having a document attribute matching a search document attribute of the at least one synonym. A processor lists the extracted document as a search result.

Abstract: A method for producing a packaging body of the present invention includes a production step of producing a metal oxide powder, and a sealing step of placing the metal oxide powder in a bag that is filled with at least one gas selected from the group consisting of a noble gas, nitrogen gas, oxygen gas, and dry air, followed by sealing the bag, the metal oxide powder being a silica powder or a magnesia powder, the bag having a moisture permeability as measured according to JIS Z 0222-1959 of 0.30 g/(m2·day) or less.

Abstract: Provided is a fluorine-based resin that has liquid-repellency and high solubility in a fluorine-based solvent, and becomes insoluble in the solvent by photocrosslinking at low light exposure intensity. A fluorine-based resin of the present invention includes: a repeating unit that is represented by Formula (1) including a photocrosslinkable group; and a repeating unit that includes a fluorine atom, where R1 represents hydrogen or methyl, L1 represents a single bond or a linking group, A represents a linking group, and R2 through R6 represent one selected from the group consisting of hydrogen, halogen, alkyl, alkyl halide, alkoxy, aryl, aryloxy, cyano, and amino.

Abstract: A spherical inorganic powder having, in a particle size frequency distribution, a first peak having a local maximum value in a particle size region from 0.1 ?m or larger to 0.6 ?m or smaller, and a second peak having a local maximum value in a particle size region from 1.5 ?m or larger to 5.0 ?m or smaller, wherein a value of a ratio of a frequency at the local maximum value of the second peak to a frequency at the local maximum value of the first peak is 1.5 or higher and lower than 10.0.

Abstract: A spherical inorganic oxide powder wherein a volume-based cumulative 50% diameter D50 is 4-55 ?m; and in a cross-section of a cured body containing an epoxy resin and the spherical inorganic oxide powder at a mass ratio of 2:1, when a total of 5000 particles with a maximum diameter of 51 ?m or larger are observed in a field of view at 100× magnification using a scanning electron microscope, a total number of air bubbles having a maximum diameter of 1 ?m or larger and smaller than 10 ?m is 40 or fewer and a total number of air bubbles having a maximum diameter of 10 ?m or larger is 30 or fewer.

Abstract: An inorganic powder having a volume-based cumulative 50% diameter D50 of 1.2-2.8 ?m, having a volume-based particle size frequency distribution that includes at least a first peak and a second peak located on a larger-particle-diameter side of the first peak, wherein the first peak has a half-width smaller than 0.5 ?m.