Abstract: A processor-implemented method includes: generating first input data comprising phase information of an input image; generating second input data in which lens position information is encoded; and determining position information of a lens corresponding to autofocus by inputting the first input data and the second input data to a neural network model.

Abstract: A method, implemented by a processor, of correcting lighting of an image includes inputting an input image to a first neural network and generating predicted lighting data corresponding to lighting of the input image and embedding data corresponding to a feature of the input image, inputting the generated predicted lighting data, the generated embedding data, and sensor data to a second neural network and generating a lighting weight corresponding to the input image, and generating correction lighting data for the input image by applying the generated lighting weight to preset basis lighting data corresponding to the input image.

Abstract: An implant removal device includes a body, a first arm, a second arm, an actuation interface, and an extraction member. The first arm is supported at the body and configured to move relative to the body between a retracted position and a skin gripping position. The second arm is supported at the body and configured to move relative to the body between the retracted position and the skin gripping position. The actuation interface is supported at the body. The actuation interface is configured to receive a first actuation input thereat to cause at least one of the first arm and the second arm to move from the skin gripping position to the retracted position. The extraction member is supported at the body and located between the first arm and the second arm.

Abstract: An implant removal device includes a body, a first arm, a second arm, an actuation interface, and an extraction member. The first arm is supported at the body and configured to move relative to the body between a retracted position and a skin gripping position. The second arm is supported at the body and configured to move relative to the body between the retracted position and the skin gripping position. The actuation interface is supported at the body. The actuation interface is configured to receive a first actuation input thereat to cause at least one of the first arm and the second arm to move from the skin gripping position to the retracted position. The extraction member is supported at the body and located between the first arm and the second arm.

Abstract: The present invention presents a scrubber burner composed of a preheating spray ring that is formed with a porous material with certain thickness, and that preheats and sprays the fuel gas in the preheating combustion space formed inside, a preheating guide ring equipped with multiple preheating guide holes that wrap the outer circumference of the aforementioned preheating spray ring, and that penetrates from the outer circumference to the inner circumference, and a preheating burner module equipped with a housing that forms a ring shaped gas channel that is separated from the outer circumference of the aforementioned preheating guide ring and through which the aforementioned fuel gas flows.

Abstract: The present disclosure relates to a method and an apparatus for federated learning of an artificial intelligence model. According to an exemplary embodiment of the present disclosure, a federated learning method of an artificial intelligence model includes: training a first local artificial intelligence model and a second local artificial intelligence model using data sets of a first client and a second client among the plurality of clients; calculating performance values for the first local artificial intelligence model and the second local artificial intelligence model by transmitting the first local artificial intelligence model to the second client and transmitting the second local artificial intelligence model to the first client; comparing the performance values to remove one of the first client and the second client; and training a global model using a client which is not removed.

Abstract: The present invention presents a burner for scrubbers equipped with the first nozzle that is in the shape of a pipe and that is equipped with the first main body, the second nozzle that is in the shape of a pipe and equipped with the second main body that is located to wrap the outer circumference of the aforementioned first main body, and the third nozzle that is in the shape of a pipe and that is equipped with the third main body located to wrap the outer circumference of the aforementioned second main body.

Abstract: An Out-Of-Distribution (OOD) detection method performed by an OOD detection device is provided. The OOD detection method includes generating a noise filter using training data based on a target distribution and using a classification model pre-trained according to the target distribution, training the generated noise filter to reduce a loss function for a first output value obtained by applying the training data passed through the generated noise filter to the pre-trained classification model, and detecting Out-Of-Distribution (OOD) for new data using a second output value obtained by applying the new data passed through the trained noise filter to the pre-trained classification model.

Abstract: A local model training method of a federated learning framework implementing training data classification is provided. In the local model training method, a client may classify training data into two categories, generate a learning mini-batch by adjusting a ratio between samples classified into the two categories and included in the mini-batch to a preset ratio, and train a learning model using the mini-batch with the adjusted sample ratio.

Abstract: A neural network model training method and apparatus are provided. The method includes a first training operation of training the neural network model with original data, the first training operation including generating a first feature map for the original data, and generating a first class activation map for the original data from the generated first feature map, and a second training operation of training the neural network model with adversarial data transformed from the original data, the second training operation including generating a second feature map for the adversarial data, generating a second class activation map for the adversarial data from the generated second feature map, and training the neural network model so that the second class activation map follows the first class activation map based on logit pairing for the first and second class activation maps.

Abstract: Provided is a method of removing, by a server, an abnormal client in federated learning. A method of removing, by a server, an abnormal client in federated learning may include receiving, from a user equipment (UE), first weight values trained in a first local model, generating a first client model based on the first weight values, validating the first client model by using a validation data set in order to determine whether the first client model is legitimate, and removing the first weight values based on the first client model not being legitimate.

Abstract: An implant removal device includes a body, a first arm, a second arm, an actuation interface, and an extraction member. The first arm is supported at the body and configured to move relative to the body between a retracted position and a skin gripping position. The second arm is supported at the body and configured to move relative to the body between the retracted position and the skin gripping position. The actuation interface is supported at the body. The actuation interface is configured to receive a first actuation input thereat to cause at least one of the first arm and the second arm to move from the skin gripping position to the retracted position. The extraction member is supported at the body and located between the first arm and the second arm.

Abstract: Provided are a method and an apparatus for generating a keyword. The method for generating a keyword in a target document includes: extracting primitive single words from the target document; generating a candidate keyword by a combination of the primitive single words; calculating cohesion score between a first word and a second word included in the candidate keyword; calculating a context score of the first word and the second word based on similarity between two contexts by determining a periphery of a word where the first word and the second word simultaneously occur in a neighboring document as one context and determining the target document as a remaining one context; and selecting a final keyword based on the cohesion score between the first word and the second word and the context score of the first word and the second word.

Abstract: Provided are a method and an apparatus for pruning based on the number of updates, the method comprising counting the number of updates of weights for each epoch and removing n weights for which the number of updates is small after training of a network is completed, wherein n represents the number of weights to be removed to satisfy a desired sparsity.

Abstract: A release actuator for a vehicle seat includes a spur gear that is rotatable in one direction or in the other direction by receiving rotational force from a motor, and has sliding grooves that are formed in both surfaces of the spur gear and have a predetermined length. A first cable guide is disposed to face one surface of the spur gear and has a protrusion placed in one of the sliding grooves. The first cable guide is connected to a first cable that withdraws a walk-in latch and releases walk-in lock when the first cable is pulled. A second cable guide is disposed to face the other surface of the spur gear, has a protrusion placed in the other sliding groove, and is connected to a second cable that withdraws a folder lock latch and releases folder lock when the second cable is pulled.

Abstract: A release actuator for a vehicle seat includes a spur gear that is rotatable in one direction or in the other direction by receiving rotational force from a motor, and has sliding grooves that are formed in both surfaces of the spur gear and have a predetermined length. A first cable guide is disposed to face one surface of the spur gear and has a protrusion placed in one of the sliding grooves. The first cable guide is connected to a first cable that withdraws a walk-in latch and releases walk-in lock when the first cable is pulled. A second cable guide is disposed to face the other surface of the spur gear, has a protrusion placed in the other sliding groove, and is connected to a second cable that withdraws a folder lock latch and releases folder lock when the second cable is pulled.

Abstract: Provided herein is an image forming apparatus, comprising: a consumable unit where a CRUM (Customer Replaceable Unit Monitoring) chip is mounted; and a main body configured to perform at least one of a first authentication and a second authentication of the consumable unit, when the consumable unit is mounted, wherein the main body comprises: a main controller for performing the first authentication according to firmware stored in the image forming apparatus; and an authentication controller for using at least one ASIC (Application Specific Integrated Circuit) to perform the second authentication of the consumable unit. Accordingly, it is possible to effectively authenticate a consumable unit even when the image forming apparatus is hacked.

Abstract: Provided are a method and an apparatus for generating a keyword. The method for generating a keyword in a target document includes: extracting primitive single words from the target document; generating a candidate keyword by a combination of the primitive single words; calculating cohesion score between a first word and a second word included in the candidate keyword; calculating a context score of the first word and the second word based on similarity between two contexts by determining a periphery of a word where the first word and the second word simultaneously occur in a neighboring document as one context and determining the target document as a remaining one context; and selecting a final keyword based on the cohesion score between the first word and the second word and the context score of the first word and the second word.

Abstract: The present invention enables a user to receive a financial service anywhere through a mobile terminal equipped with a UIM (User Identification Module) electronic card. In the present invention, a user enters his or her password to a mobile terminal with a UIM card including subscriber telephone number, finance, authorization, and personal information, then, if the entered password is correct, authorization is processed with a remote authorizing server based on the authorization information. After authorization, user's requesting service, e.g., payment service, transaction particulars inquiry service, prepaid card recharging service is conducted through a mobile network.

Abstract: An image scanning apparatus and method are provided. The image scanning apparatus includes a scanner configured to scan a document to generate a scan image, a driver configured to move the document or the scanner, a position sensor configured to, in response to the scanner or the document passing a preset position, generate a position interruption, and a controller configured to control the driver to move the scanner or the document according to a scan command. The scanner scans the document without control of the controller after the position interruption is generated and then a preset time elapses.