Abstract: A multilayer ceramic electronic component includes: a ceramic body including an active portion having dielectric layers and first and second internal electrodes and first and second cover portions disposed on opposite surfaces of the active portion in a stacking direction, respectively; wherein when a region of the cover portion in contact with the first or second internal electrode is an inner region of the cover portion and a region of the active portion in contact with the inner region of the cover portion is an outer region of the active portion, 1.00<XA/XB?1.04 in which XA/XB is a ratio of a molar ratio (XA) of barium (Ba) to titanium (Ti) in the inner region of the cover portion to a molar ratio (XB) of barium (Ba) to titanium (Ti) in the outer region of the active portion.

Abstract: The present disclosure describes a system, method, and computer program for predicting sentiment labels for audio speech utterances using an audio speech sentiment classifier pretrained with pseudo sentiment labels. A speech sentiment classifier for audio speech (“a speech sentiment classifier”) is pretrained in an unsupervised manner by leveraging a pseudo labeler previously trained to predict sentiments for text. Specifically, a text-trained pseudo labeler is used to autogenerate pseudo sentiment labels for the audio speech utterances using transcriptions of the utterances, and the speech sentiment classifier is trained to predict the pseudo sentiment labels given corresponding embeddings of the audio speech utterances. The speech sentiment classifier is then subsequently fine tuned using a sentiment-annotated dataset of audio speech utterances, which may be significantly smaller than the unannotated dataset used in the unsupervised pretraining phase.

Abstract: For any application that processes speech, improving the quality of the feature vectors may improve the quality of the speech application. The quality of feature vectors may be improved by modifying a neural network architecture for computing feature vectors to allocate computational resources where they are more effective for learning and computing the feature vectors. Contextual feature vectors may be computed from feature vectors by using a parameterized downsampling operation that decreases a vector sequence rate, processing the downsampled vectors with a neural network, and using a parameterized upsampling operation that increases a vector sequence rate. For example, parameterized downsampling may decrease a vector sequence rate by a factor of two, a neural may require fewer computational resources since it operates with a lower vector sequence rate, and parameterized upsampling may then increase the vector sequence rate by a factor of two.

Abstract: A vehicle braking control method is provided. When a driver intends to drive a vehicle after the delivery of the vehicle or a factory mechanic intends to test the vehicle before the delivery of the vehicle after the engine is turned on even when a warning light is turned on due to the insufficiency of the brake fluid, a warning signal indicating that a level sensor is malfunctioning is generated using an instrument cluster, or driving torque of the engine is limited while a warning phrase indicating the insufficiency of the brake fluid is displayed using the cluster. Therefore, the vehicle may travel at a minimum speed. Thus, the driver is enabled to drive the vehicle to a safe place. Accordingly, a secondary accident is prevented and a subsequent maintenance operation is easily performed.

Abstract: The amount of future context used in a speech processing application allows for tradeoffs between performance and the delay in providing results to users. Existing speech processing applications may be trained with a specified future context size and perform poorly when used in production with a different future context size. A speech processing application trained using a stochastic future context allows a trained neural network to be used in production with different amounts of future context. During an update step in training, a future-context size may be sampled from a probability distribution, used to mask a neural network, and compute an output of the masked neural network. The output may then be used to compute a loss value and update parameters of the neural network. The trained neural network may then be used in production with different amounts of future context to provide greater flexibility for production speech processing applications.

Abstract: An embodiment provides a semiconductor device comprising: a substrate; a semiconductor structure disposed on the substrate and including a first conductive semiconductor layer, a second conductive semiconductor layer, and an activation layer disposed between the first conductive semiconductor layer and the second conductive semiconductor layer; a bonding layer disposed between the semiconductor structure and the substrate; a cover layer disposed between the bonding layer and the semiconductor structure; and an electrode pad disposed on the cover layer and spaced apart from the semiconductor structure, wherein: the semiconductor structure further comprises a stepped portion at which the lateral surface of the second conductive semiconductor layer, the lateral surface of the activation layer, and the lower surface of the first conductive semiconductor layer are exposed; the stepped portion is disposed at the outer portion of the semiconductor structure; and the cover layer is disposed to extend from a region ve

Abstract: A connection structure of a stator of a drive motor is configured to allow coils to be wound in a plurality of slots provided in a stator core and to connect the coils withdrawn from the slots. The coils are wound in the slots to form first-type structures and second-type structures configured such that withdrawal directions of three-phase (U-, W- and V-phase) withdrawal lines and N-phase withdrawal lines withdrawn from the slots of the first-type structures are opposite to withdrawal directions of three-phase (U-, W- and V-phase) withdrawal lines and N-phase withdrawal lines withdrawn from the slots of the second-type structures. The first-type structures and the second-type structures are disposed symmetrically to each other with respect to a reference line formed to divide the slots in half.

Abstract: An autonomous cleaner includes a body having a first housing formed at a front and a second housing formed at a rear of the first housing; a brush unit installed at the first housing and configured to sweep and collect dust from a floor; a dust collecting unit installed at the second housing and configured to store the dust inlet into the brush unit; a driving unit to drive the body and coupled to the second housing to be positioned at a lateral side of the dust collecting unit; and a power unit installed at the second housing and coupled to be positioned at a rear of the dust collecting unit. The miniaturization of the autonomous cleaner may be provided while at the same time the capacity of a dust collecting container and the capacity of a battery are increased.

Abstract: A display apparatus comprising a driving circuit that includes a storage capacitor and at least one thin film transistor, and a light-emitting device that includes a region disposed outside the driving circuit and a region overlapping with the storage capacitor, each of capacitor electrodes of the storage capacitor being a transparent electrode having a relatively higher transmittance, so that the quality of the image realized on an outer surface of a device substrate which supports the driving circuit and the light-emitting device of each pixel area can be improved.

Abstract: Provided herein is a memory device and a method of operating the same. The memory device may include a plurality of memory cells, a peripheral circuit, and a control logic. The peripheral circuit may be configured to perform a plurality of program loops, each including a program pulse apply operation and a program verify operation, on selected memory cells of the plurality of memory cells. The control logic may be configured to control, in response to a suspend command, the peripheral circuit to suspend an n-th program loop of the plurality of program loops, where n is a natural number of 1 or more, and configured to control, in response to a resume command, the peripheral circuit to resume the suspended n-th program loop after performing a recovery pulse apply operation compensating for charges detrapped from a channel area of the selected memory cells.

Abstract: A multilayer ceramic electronic component includes: a ceramic body including an active portion having dielectric layers and first and second internal electrodes and first and second cover portions disposed on opposite surfaces of the active portion in a stacking direction, respectively; wherein when a region of the cover portion in contact with the first or second internal electrode is an inner region of the cover portion and a region of the active portion in contact with the inner region of the cover portion is an outer region of the active portion, 1.00<XA/XB<1.04 in which XA/XB is a ratio of a molar ratio (XA) of barium (Ba) to titanium (Ti) in the inner region of the cover portion to a molar ratio (XB) of barium (Ba) to titanium (Ti) in the outer region of the active portion.

Abstract: A cleaner and a method for controlling the same. The cleaner includes a brush module comprising a brush configured to scatter foreign substances on a floor, and a brush motor configured to rotate the brush. The cleaner includes a suction module configured to suction the foreign substances; a sensor configured to detect a load applied to the brush; and a controller configured to control suction force of the suction module based on the load detected by the sensor.

Abstract: A vehicle braking control method is provided. When a driver intends to drive a vehicle after the delivery of the vehicle or a factory mechanic intends to test the vehicle before the delivery of the vehicle after the engine is turned on even when a warning light is turned on due to the insufficiency of the brake fluid, a warning signal indicating that a level sensor is malfunctioning is generated using an instrument cluster, or driving torque of the engine is limited while a warning phrase indicating the insufficiency of the brake fluid is displayed using the cluster. Therefore, the vehicle may travel at a minimum speed. Thus, the driver is enabled to drive the vehicle to a safe place. Accordingly, a secondary accident is prevented and a subsequent maintenance operation is easily performed.

Abstract: An embodiment relates to a light emitting module, a flash module, and a terminal including the same. The light emitting module according to an embodiment comprises: a semiconductor layer; a phosphor layer arranged on one surface of the semiconductor layer; a plurality of light emitting chips including a plurality of electrodes arranged on a surface facing one surface of the semiconductor layer; a first partition arranged at one side of the plurality of light emitting chips, and a second partition arranged at the other side of the plurality of light emitting chips so as to face the first partition; and an opaque molding part, which encompasses the plurality of light emitting chips such that the upper surface of the phosphor layer and the bottom surfaces of the plurality of electrodes are exposed to the outside, and is arranged on the inner side of the first and second partitions.

Abstract: A cleaner and a method for controlling the same. The cleaner includes a brush module comprising a brush configured to scatter foreign substances on a floor, and a brush motor configured to rotate the brush. The cleaner includes a suction module configured to suction the foreign substances; a sensor configured to detect a load applied to the brush; and a controller configured to control suction force of the suction module based on the load detected by the sensor.

Abstract: A stator for a drive motor includes: a segmented stator core including a plurality of mounting portions; a bobbin equipped on each mounting portion of the plurality of the mounting portions and having a coil wound thereon; a first casing and a second casing coupled to the bobbin, respectively, at opposite sides of the bobbin with respect to the segmented stator core, and configured to enclose the coil; a passage formed in one of the first casing or the second casing, and configured to allow fluid communication with the inside of the first casing and the second casing; and a terminal portion integrally formed in a remaining one of the first casing or the second casing and configured to allow connection of the coil.

Abstract: A robot cleaner and a charging device capable of determining whether contact is made between charging terminals of the charging device and the robot cleaner are provided. The charging device may include a charging circuit including at least one terminal having at least a portion exposed to the outside, at least one object sensor to detect at least one identification object arranged in a robot cleaner, the at least one object sensor being arranged separately from the at least one terminal configured, and at least one processor configured to control the charging circuit to apply a voltage to the at least one terminal in response to the at least one object sensor detecting the at least one identification object.

Abstract: A robot cleaner and a method for controlling the same are disclosed, which perform efficient cleaning by controlling a suction force or traveling route of the robot cleaner. The robot cleaner detects load applied to wheels of the robot cleaner so as to increase reliability and accuracy in floor state decision, thereby recognizing a floor state. The robot cleaner recognizes a floor state by combining load applied to wheels of the robot cleaner, load applied to brushes, and acceleration information of the robot cleaner with one another in a complementary manner, thereby increasing accuracy in floor state decision.

Abstract: The present disclosure relates to a system, method, and computer program for creating a complete transcription of an audio recording from separately transcribed redacted and unredacted words. The system receives an original audio recording and redacts a plurality of words from the original audio recording to obtain a modified audio recording. The modified audio recording is outputted to a first transcription service. Audio clips of the redacted words from the original audio recording are extracted using word-level timestamps for the redacted words. The extracted audio clips are outputted to a second transcription service. The system receives a transcription of the modified audio recording from the first transcription service and transcriptions of the extracted audio clips from the second transcription service.

Abstract: An embodiment relates to a light emitting module, a flash module, and a terminal including the same. The light emitting module according to an embodiment comprises: a semiconductor layer; a phosphor layer arranged on one surface of the semiconductor layer; a plurality of light emitting chips including a plurality of electrodes arranged on a surface facing one surface of the semiconductor layer; a first partition arranged at one side of the plurality of light emitting chips, and a second partition arranged at the other side of the plurality of light emitting chips so as to face the first partition; and an opaque molding part, which encompasses the plurality of light emitting chips such that the upper surface of the phosphor layer and the bottom surfaces of the plurality of electrodes are exposed to the outside, and is arranged on the inner side of the first and second partitions.