Abstract: A method of processing speech includes: obtaining a speech input; obtaining an instruction related to the speech input; obtaining a speech representation corresponding to the speech input; obtaining an adapter that includes speech information by fusing a pre-trained adapter with the speech representation; and obtaining a response corresponding to the instruction by inputting both the adapter that includes the speech information and the instruction to a language model, the language model generating the response based on the adapter that includes the speech model and the speech information.

Abstract: A method of manufacturing a color-routing element, may include: generating an initial pattern; performing blurring on the initial pattern to generate a reference pattern; performing edge detection on the reference pattern to generate at least one comparison pattern reflecting a process error; performing a simulation to obtain at least one color-routing figure of merit based on the reference pattern and the at least one comparison pattern; updating the initial pattern based on a calculation result of the at least one color-routing figure of merit; generating the updated initial pattern as a target pattern of the color-routing element; and manufacturing the color-routing element based on the target pattern.

Abstract: A carrageenan derivative according to the present disclosure can quickly and easily detect and label the presence of inflammation-related macrophages and, particularly, can accurately and easily identify M1-type macrophages highly associated with early inflammation. In addition, the carrageenan derivative has advantages in that it can be produced easily, environmental pollution is minimized because a small amount of organic solvent is used, and initial cost is low.

Abstract: An electronic device includes: a display layer in which an active area and a peripheral area adjacent to the active area are defined, the display layer including a common electrode, a sensor layer disposed on the display layer and including a plurality of sensing electrodes and an antenna pattern, and a cover layer spaced apart from the sensor layer with the display layer therebetween, the cover layer having a conductivity. The peripheral area includes a first area and a second area spaced apart from the active area with the first area therebetween, and the common electrode is disposed in the active area and the first area and is spaced apart from the second area. When viewed in a plan view, a portion of the antenna pattern does not overlap the common electrode and overlaps the cover layer.

Abstract: A remote computing apparatus communicating with a local computing apparatus is disclosed. The apparatus includes a communication interface, a storage device, a memory in which a driver of the storage device is executed, and a processor that stores data and metadata associated with the data in the storage device. The driver receives input/output commands for the storage device from the local computing apparatus through the communication interface, the processor processes the received input/output commands for the storage device using a plurality of queues, and the driver provides a result of processing the input/output commands to the local computing apparatus through the communication interface. By providing this apparatus, latency of remote memory operations can be reduced.

Abstract: The gait assistance mechanism according to one example of the present invention may comprise one or more tendons connecting an upper limb, a trunk, and a lower limb of a wearer and a fixing part fixing the one or more tendons to the body of the wearer, and the one or more tendons may comprise a driving tendon connecting the lower limb and the upper limb on the same side of the wearer.

Abstract: The present disclosure relates to an integrated control system for heat pump to optimize energy of solar power generation building, Including a computer comprising a processor and a memory within which a code for execution by the processor is stored, and the computer is configured to collect real-time data measured from at least any one of the heat pump of the solar power generation building, an external environment, a room and a photovoltaic power generation system, where interactions occur in a dynamic behavior of the heat pump; input the real-time data of before pre-set time into multiple dynamic behavior prediction models to derive an indoor air temperature prediction value, heat pump electricity consumption prediction value, and photovoltaic power generation prediction value, of after pre-set time; and search for an optimal control variable of the heat pump using the derived indoor air temperature prediction value, heat pump electricity consumption prediction value, and photovoltaic power generation pred

Abstract: The present invention relates to a composition for detecting hydrogen sulfide or measuring a concentration of hydrogen sulfide, and a composition comprising same as an effective ingredient for diagnosing or imaging in vivo inflammation, tissues having hypoxic damage, or cancer.

Abstract: Disclosed is a perovskite light-emitting device with reduced defects in a perovskite thin film. The passivation layer in the perovskite light-emitting device is formed on the upper part of the perovskite thin film to eliminate defects in the perovskite nanocrystalline particles and resolve charge imbalance in the device, thereby improving maximum efficiency and maximum luminance of the light-emitting device.

Abstract: The present invention relates to a medium composition for culture of extraembryonic endoderm stem cells, and more specifically, to a medium composition for in vitro culture of extraembryonic endoderm stem cells.

Abstract: An LED device includes a light emission layer including a first semiconductor layer including a first surface and a second surface facing the first surface, the second surface having an area larger than an area of the first surface, an active layer on the first surface, and a second semiconductor layer on the active layer, an insulating layer on the first surface, the insulating layer defining an open region of the first semiconductor layer, a first electrode that contacts the first semiconductor layer via the open region, and a second electrode on the second semiconductor layer and contacting the second semiconductor layer.

Abstract: A retinal projection display device is provided. The retinal projection display device includes a light source configured to emit light, a spatial light modulator configured to generate diffracted light by diffracting the emitted light, a holographic optical element configured to reflect the diffracted light by duplicating the diffracted light into a plurality of complex wavefronts, and a field lens configured to focus the plurality of complex wavefronts to a plurality of respective focal points in an eye-box, wherein the plurality of complex wavefronts overlap each other.

Abstract: A compound represented by chemical formula 1, a salt, a stereoisomer, a solvent or a hydrate thereof, an antibacterial composition including the compound, and uses thereof are disclosed. The antibacterial composition is suitable for inhibiting bacterial infections, preventing, alleviating or treating bacterial infectious disease, and for preventing, alleviating or treating inflammation caused by a bacterial infection. The compound specifically binds to p62 protein so as to activate p62, thereby increasing the activity of autophagy, which is inhibited by a bacterial infection, and thus has antibacterial activity against various bacteria, can inhibit bacterial infectivity and has the excellent effect of enabling prevention, alleviation or treatment of bacterial infectious disease and inflammation caused by an infection.

Abstract: A scheduler, a method of operating the scheduler, and an accelerator apparatus including the scheduler are disclosed. A method of operating a scheduler to perform scheduling on models to be executed in an accelerator, the method includes receiving at least one execution request for a first model and a second model that are executed independently from each other in the accelerator, and performing layer-unit scheduling on the first model and the second model based on workload characteristics of the first model and the second model.

Abstract: The present disclosure relates to a flame transfer function measurement system for prediction and reduction of combustion instability.

Abstract: The present application relates to a cosmetic composition comprising resveratrone, for improving skin health.

Abstract: A neural network device and an action selecting method using the same, which select an action corresponding to a current state on the basis of a value return. A method of selecting, executed by at least one processor, an action on the basis of deep learning includes receiving a current state as an input, calculating a value distribution corresponding to each of a plurality of actions to be performed on the current state, and selecting an optimal action from among the plurality of actions based on the value distribution, wherein the value distribution includes at least one Gaussian graph following a Gaussian distribution.

Abstract: An aircraft flight scheduling apparatus according to an embodiment of the present disclosure includes a database configured to manage an arrival time and a departure time for each aircraft at each airport, aircraft flight data including slot information assigned to each aircraft at each airport, a ground delay program (GDP) information issued by a control center of each airport, a scenario for an expected aircraft flight according to generation of the GDP, and an objective function for determining resetting of an aircraft flight schedule according to the generation of the GDP, a memory for storing an aircraft flight scheduling program, and a processor configured to execute the aircraft flight scheduling program.

Abstract: The present disclosure relates to a method for more readily preparing an oxime derivative under mild conditions by using a nitrogen oxide as a NO supply source.

Abstract: Provided are a cathode for a secondary battery, a manufacturing method therefor, and a lithium secondary battery comprising same, the cathode comprising: a cathode current collector; a first layer which is arranged on at least one surface of the cathode current collector and which comprises a first cathode active material; and a second layer which is arranged on the first layer and which comprises a second cathode active material, wherein the first cathode active material is represented by the following chemical formula 1, the second cathode active material is represented by the following chemical formula 1 or 2, and the first cathode active material and the second cathode active material are different from each other. LixNiyM1?yO2??<Chemical formula 1> Li?Co?M?1??O2??<Chemical formula 2> See the Detailed Description of the Invention for the definitions of x, y, a, p, M and M? in the formulas.