Abstract: One embodiment of the present disclosure provides a magnetic fluid seal including a rotating shaft, a pole piece disposed to surround the rotating shaft, a plurality of trenches formed on any one of the rotating shaft and the pole piece, a magnetic fluid provided between the other one of the rotating shaft and the pole piece and at least one of ends of the plurality of trenches, a blocking member provided in a space between the plurality of trenches, and a permanent magnet configured to apply a magnetic force to the magnetic fluid.

Abstract: A substrate analysis apparatus is provided. The substrate analysis includes: an interlayer conveying module configured to transport a first FOUP; an exchange module which is connected to the interlayer conveying module, and configured to transfer a wafer from the first FOUP to a second FOUP; a pre-processing module configured to form a test wafer piece using the wafer inside the second FOUP; an analysis module configured to analyze the test wafer piece; and a transfer rail configured to transport the second FOUP containing the wafer and a tray containing the test wafer piece. The wafer includes a first identifier indicating information corresponding to the wafer, the test wafer piece includes a second identifier indicating information generated by the pre-processing module which corresponds to the test wafer piece, and the analysis module is configured to analyze the first identifier and the second identifier in connection with each other.

Abstract: A method of displaying an image in a display device may include determining the degree of deterioration of pixels included in a display unit based on image data of a current frame image, determining a shift route of the current frame image so as to correspond to the determined degree of deterioration. The first image data is corrected to second image data so that the current frame image is shifted along the shift route.

Abstract: There is provide a method for manufacturing analytical semiconductor samples by using an apparatus for manufacturing analytical semiconductor samples, which minimizes a feedback time by manufacturing a viewing surface that is environment-friendly and has a large area. The method comprising mounting the analytical semiconductor samples to a holder; discharging deionized (DI) water to an upper surface of a polishing plate through a DI water nozzle; grinding the analytical semiconductor samples with the upper surface of the polishing plat; determining whether a desired viewing surface of the analytical semiconductor samples has been acquired after the grinding of the analytical semiconductor samples; and transferring the analytical semiconductor samples to analyze the viewing surface of the ground analytical semiconductor samples based on a determination that the desired viewing surface of the analytical semiconductor samples has been acquired.

Abstract: The present invention relates to a fusion protein comprising a SARS-CoV-2-derived receptor-binding domain and a nucleocapsid protein, and the use thereof. The fusion protein comprising a coronavirus-derived receptor-binding domain and a nucleocapsid protein is highly applicable to a multivalent vaccine composition having greatly improved in-vivo half-life and remarkably superior efficacy compared to an immunogenic composition comprising only a receptor-binding domain. In particular, the fusion protein can greatly improve the titer of the coronavirus-specific antibody formation and T-cell immune response, and is thus useful for the prevention and treatment of coronaviruses comprising SARS-CoV-2.

Abstract: Systems and methods are provided for dynamically adjusting payout to deliverers, comprising receiving a plurality of delivery orders, wherein each of the plurality of delivery orders is associated with a delivery address in a geographic region, and wherein the geographic region is associated with a plurality of configurations; determining a base fee for fulfilling the plurality of delivery orders in the geographic region based on the plurality of configurations associated with the geographic region; adjusting, using a pricing algorithm, the base fee for fulfilling the plurality of delivery orders based on the determined configurations; receiving one or more features associated with the plurality of delivery orders; and adjusting the adjusted base fee based on the received one or more features, wherein the pricing algorithm used to calculate the base fee is selected among a plurality of pricing algorithms based on the geographic region.

Abstract: Systems and methods are provided for integrating and deduplicating products using Al. One method comprises receiving at least one request to register a first product; searching at least one data store for a second product; tagging, using a machine learning model, at least one keyword from product information associated with the first product and tagging at least one keyword from product information associated with the second product; determining, using the machine learning model, a match score between the first product and the second product; when the match score is above a first predetermined threshold, determining, using the machine learning model, that the first product is identical to the second product; and when the match score is below a first predetermined threshold, determining, using the machine learning model, that the first product is not the second product.

Abstract: Systems and methods are provided for categorizing products using AI. One method comprises retrieving initial training data including products associated with one or more categories; pre-processing the initial training data to generate synthesized training data; generating a hierarchical model using the synthesized training data, the hierarchical model containing at least two layers of nodes below a root node; receiving information associated with a first uncategorized product; and receiving a request to predict a set of N categories with the highest N total probability scores.

Abstract: Disclosed are systems and methods for multi-point destination arrival time analysis. In one aspect, the system may include: a memory storing instructions; and at least one processor configured to execute the instructions to: receive a plurality of durations associated with at least one orderable item and associated metadata, the durations based on past history of producing the orderable items; store the plurality of durations associated with the at least one orderable item in a data store; retrieve, from the data store, the stored plurality of durations associated with the at least one item; receive an estimated preparation time for the at least one orderable item from a machine-learning algorithm; determine, based on the estimated preparation time, an assigned delivery worker associated with a second external system; and forward a request to the second external system to fulfill a delivery of the at least one orderable item.

Abstract: Systems and methods are provided for categorizing products using Al. One method comprises retrieving initial training data including products associated with one or more categories; pre-processing the initial training data to generate synthesized training data; generating a hierarchical model using the synthesized training data, the hierarchical model containing at least two layers of nodes below a root node; receiving information associated with a first uncategorized product; and receiving a request to predict a set of N categories with the highest N total probability scores.

Abstract: The present invention provides a composition for cleaning a mask comprising a) 0.2 to 10% by weight of two or more alkali compounds selected from alkali hydroxides and alkali carbonates, b) 0.1 to 8% by weight of nitrate, c) 1 to 8% by weight of an oxidizer, and d) residual water, wherein the pH is 11.5 to 14. According to the present invention, a composition for cleaning a mask and a method for cleaning a mask using the same, which can remove only the film deposited on the mask in a short time without damaging the mask, was provided.

Abstract: A semiconductor device and a manufacturing method thereof are provided. The semiconductor device includes: a deep trench in a substrate; a sidewall insulating film on a side surface of the deep trench; an interlayer insulating film on the sidewall insulating film; and an air gap in the interlayer insulating film.

Abstract: A semiconductor device and a manufacturing method thereof are provided. The semiconductor device includes: a deep trench in a substrate; a sidewall insulating film on a side surface of the deep trench; an interlayer insulating film on the sidewall insulating film; and an air gap in the interlayer insulating film.

Abstract: A semiconductor device and a manufacturing method thereof are provided. The semiconductor device includes: a deep trench in a substrate; a sidewall insulating film on a side surface of the deep trench; an interlayer insulating film on the sidewall insulating film; and an air gap in the interlayer insulating film.

Abstract: A semiconductor device and a manufacturing method thereof are provided. The semiconductor device includes: a deep trench in a substrate; a sidewall insulating film on a side surface of the deep trench; an interlayer insulating film on the sidewall insulating film; and an air gap in the interlayer insulating film.

Abstract: An apparatus and method for extending a pronunciation dictionary for speech recognition are provided. The apparatus and the method may segment speech information of an input utterance into at least one phoneme, collect segmentation information of the at least one segmented phoneme, analyze a pronunciation variation of the at least one segmented phoneme based on the collected segmentation information, and select a substitutable phoneme group for the at least one phoneme where the pronunciation variation occurs, and extend the pronunciation dictionary.

Abstract: Disclosed is a DPDT RF switch. The DPDT RF switch includes: first to fourth transmission lines for forming first to fourth ports, respectively; and first to fourth slot line pattern sections. The first slot line pattern section includes: a first slot line; and a first switching device for blocking signal transfer by short-circuiting a gap of a slot line. The third slot line pattern section includes: a third slot line; and a third switching device for blocking signal transfer by short-circuiting a gap of a slot line. The second slot line pattern section includes: a first loop-shaped slot line; a second slot line; and a second switching device for blocking signal transfer by short-circuiting a gap of a slot line. The fourth slot line pattern section includes: a second loop-shaped slot line; a fourth slot line; and a fourth switching device for blocking signal transfer by short-circuiting a gap of a slot line.

Abstract: A speech recognition apparatus and method for real-time speaker adaptation are provided. The speech recognition apparatus may estimate a pitch of a speech section from an inputted speech signal, extract a speech feature for speech recognition based on the estimated pitch, and perform speech recognition with respect to the speech signal based on the speech feature. The speech recognition apparatus may be adaptively normalized depending on a speaker. Thus, the speech recognition apparatus may extract a speech feature for speech recognition, and may improve a performance of speech recognition based on the extracted speech feature.

Abstract: An RF switch includes first, second and third transmission lines for forming ports, respectively, and first, second and third slot line pattern portions connected to one another, for transferring signals to the first, second and third transmission lines, respectively. The first slot line pattern portion has a slot line pattern for transferring a signal received from the first transmission line to a connection point with the other slot line pattern portions, and a switching circuit for shorting the gap of a corresponding slot line and thus blocking the signal transfer. The second slot line pattern portion includes a loop slot line formed by a first and a second half loop slot line, a second sub-slot line for transferring a signal received from the connection point to the second transmission line through the loop slot line, and a switching circuit for shorting the gap of a corresponding slot line.

Abstract: An apparatus and method for extending a pronunciation dictionary for speech recognition are provided. The apparatus and the method may segment speech information of an input utterance into at least one phoneme, collect segmentation information of the at least one segmented phoneme, analyze a pronunciation variation of the at least one segmented phoneme based on the collected segmentation information, and select a substitutable phoneme group for the at least one phoneme where the pronunciation variation occurs, and extend the pronunciation dictionary.