Abstract: According to one embodiment, a learning system includes a plurality of local devices and a server. Each of the local devices includes a processor. The processor of the local device selects a first parameter set from a plurality of parameters related to the local model, and transmits the first parameter set to the server. At least one of the local devices is different from other local devices in a size of the local model in accordance with a resolution of input data. The server comprises a processor. The processor of the server integrates first parameter sets acquired from the local devices and update a global model. The processor of the server transmits the second parameter set to a local device that has transmitted the corresponding first parameter set.

Abstract: A learning device according to an embodiment includes one or more hardware processors configured to function as a generation unit, an inference unit, and a training unit. The generation unit generates input data with which an error between a value output from each of one or more target nodes and a preset aimed value is equal to or less than a preset value, the target nodes being in a target layer of a plurality of layers included in a first neural network. The inference unit causes the input data to propagate in a forward direction of the first neural network to generate output data. The training unit trains a second neural network differing from the first neural network by using training data including a set of the input data and the output data.

Abstract: According to one embodiment, a learning apparatus includes a processing circuit. The processing circuit acquires a first training condition and a first model trained in accordance with the first training condition, sets a second training condition used to reduce a model size of the first model, different from the first training condition, in accordance with the second training condition and based on the first model, trains a second model whose model size is smaller than that of the first model, and in accordance with a third training condition that is not the same as the second training condition and complies with the first training condition, trains a third model based on the second model.

Abstract: An image processing device includes a division circuit that divides an input image and outputs a plurality of divided images, a first processor that performs computation of an object detection model and acquires attribute information including an attribute value of an object included in each of the divided images and a first quadrangular frame surrounding the object as first metadata, a scaling circuit that output an overall image obtained by shrinking the input image, a second processor that performs computation of the object detection model and acquires attribute information including an attribute value of an object included in the overall image and a second quadrangular frame as second metadata, and a third processor that generates third metadata of the input image by combining pieces of the attribute information of the second metadata and pieces of attribute information that are not held in common by the first and second metadata.

Abstract: According to one embodiment, a machine learning apparatus includes a processing circuit. The processing circuit trains a first learning: parameter of an extraction layer configured to extract feature data of the input data, based on a plurality of training data. The processing circuit trains a second learning parameter of a reconstruction layer configured to generate reconstructed data of the input data, based on a plurality of training feature data obtained by applying the trained extraction layer to the plurality of training data. The second learning parameter represents representative vectors as many as a dimension count of the feature data. The representative vectors as many as the dimension count are based on a weighted sum of the plurality of training data.

Abstract: An embodiment is a data sequence correction method. The data sequence correction method including temporarily saving data with sequence information imparted thereto in a ring buffer, the ring buffer having a predetermined number of storage regions corresponding to the sequence information, and being provided with a monitoring section made up of one, or two or more consecutive sequence numbers, and an acceptance section in which a start or a second sequence number of the monitoring section is a start sequence number, and the sequence number ahead by a count of storage regions of the ring buffer including the start of the monitoring section is an end sequence number.

Abstract: A traffic monitoring apparatus includes: a header analysis circuit configured to acquire one or more identifiers from a header of a received packet; a rule registration circuit configured to convert a rule table including rules in which one or more rule elements are registered for each of the rules into a predetermined format and register the rule table in a rule matching circuit; and the rule matching circuit configured to search for rules to be matched with the acquired identifiers.

Abstract: A state detection circuit compares a target connection with an immediate previous connection based on an identifier of an incoming packet and detects a post-transitional state of the target connection based on a control flag of the incoming packet and on a pre-transitional state of the target connection detected just before if the target connection is identical with the immediate previous connection; and a connection counting circuit increments or decrements the number of target connections only when the detected post-transitional state indicates a start or end of the target connection.

Abstract: A network load balancing apparatus has a data buffer for each communication path of a received packet's transfer destinations, calculates a first hash value using a field value contained in the packet, determines, based on the field value of the packet or the first hash value, a communication path of a transfer destination of the packet subject to external transfer control for transmission to a predetermined external server, determines, based on the first hash value, a communication path of a transfer destination of the packet to be subject to priority control, determines, based on a second hash value based on the first hash value, a communication path of a transfer destination of the packet to be subject to load balancing control, to match a preset load balancing situation of the data buffer, and transmits the packet to a data buffer corresponding to the communication path of the transfer destination.

Abstract: According to one embodiment, a learning apparatus includes a processor. The processor determines, based on a data resolution of subject data obtained at a subject device, a plurality of data resolutions that differ from one another within a range covering the data resolution of the subject data, the data resolutions each indicating a corresponding amount of information per unit. The processor trains a scalable network with training samples corresponding to each of the plurality of data resolutions, the scalable network being a neural network adapted to change a data resolution of input data.

Abstract: A providing apparatus according to an embodiment of the present disclosure includes a memory and a hardware processor coupled to the memory. The hardware processor is configured to: store, in the memory, a first machine learning model capable of changing an amount of calculation of a model of a neural network; acquire device information; set, based on the device information, extraction conditions representing conditions for extracting second machine learning models from the first machine learning model; extract the second machine learning models from the first machine learning model based on the extraction conditions; and provide the second machine learning models to a device specified by the device information.

Abstract: A burst traffic detection device includes a packet receiver configured to receive packets from a network, a flow specification device configured to specify, in accordance with header information of the packets, flow rules, a flow information storage device configured to store flow information of the specified flow rules, a statistical information storage device configured to store statistical information including the total number of packets for each flow rule and/or the total number of bytes for each flow rule, a burst detection device configured to detect the occurrence of burst traffic in accordance with the statistical information, and a detection count storage device configured to store the number of times of the occurrence of burst traffic.

Abstract: A data sequence correction method for temporarily saving data with sequence information in a ring buffer and performing sequence correction is provided. The ring buffer includes a number of storage regions, a monitoring section having one or more continuous sequence numbers, and an acceptance section having a first or second sequence number of the monitoring section as a start sequence number and a sequence number immediately preceding the start sequence number of the monitoring section as an end sequence number. The method includes, when a value determined based on a remainder obtained by dividing a sequence number of received data by the number of storage regions is inside the acceptance section, writing the received data in a position of the storage region corresponding to the determined value, and when data are written in the entire monitoring section, reading out all the data in the monitoring section.

Abstract: An embodiment packet capture device comprises: a packet receiver configured to receive a packet from a network; a packet retainer configured to store the received packet in a memory to temporarily retain the received packet; a failure detector configured to determine a communication failure is present in the network; a capture controller configured to determine an operation stop address such that retention of packets from the network in time periods before and after a detection time point of the communication failure is ensured when the communication failure is detected by the failure detector; and a capture data generator configured to output the packet stored in the memory as capture data when a storage destination address of the packet stored in the memory has reached the operation stop address or when at least a predetermined waiting time period has elapsed from the detection time point of the communication failure.

Abstract: An access control unit includes packet buffers provided for each of users, a packet identification unit that stores received packets in a corresponding packet buffer, a scheduling unit that decides a packet buffer to be the object of transfer, a transfer control unit that, in a case that updating of reference data can be performed at an application processing circuit, and also the packet buffer decided by the scheduling unit is different from the current packet buffer that is the object of transfer, updates to reference data corresponding to the packet buffer decided by the scheduling unit, and a buffer selection unit that connects the packet buffers decided to be the object of transfer to the packet transfer unit when updating of reference data is completed.

Abstract: A magnetic resonance imaging apparatus according to an embodiment includes an MRI system and a processing circuitry. The MRI system includes a receiving coil to receive a magnetic resonance signal. The processing circuitry is configured to generate an image based on the magnetic resonance signal, the image including a plurality of pixels; calculate a feature value corresponding to a signal value of the pixel; correct the feature values based on a sensitivity of the receiving coil; and reduce noise in the image based on distribution of the corrected feature values.

Abstract: A photon counting X-ray CT apparatus according to an embodiment includes: data acquiring circuitry, and processing circuitry. The data acquiring circuitry is configured to allocate energy measured by signals output from a photon counting detector in response to incidence of X-ray photons to any of a plurality of first energy bins so as to acquire a first data group as count data of each of the first energy bins. The processing circuitry is configured to determine a plurality of second energy bins obtained by grouping the first energy bins in accordance with a decomposition target material that is a material to be decomposed in a imaging region, allocate the first data group to any of the second energy bins so as to generate a second data group, and use the second data group to generate an image representing a distribution of the decomposition target material.

Abstract: According to one embodiment, a learning apparatus includes a processing circuit. The processing circuit acquires first sequence data representing transition of inference performance according to a training progress of a first model trained in accordance with a first training parameter value concerning a specific training condition. The processing circuit performs iterative learning of a second model in accordance with a second training parameter value concerning the specific training condition and changes the second training parameter value based on the inference performance of the second model and the first sequence data in a training process of the second model.

Abstract: A gap formed in a front portion of a door is easily and reliably filled so that occurrence of noise is reduced. A frame front side portion retaining a glass run front side portion being inserted is provided in a front portion of a window frame. A front side of an outer panel is provided with a mirror base to which a mirror base cover is attached. A rear edge of the mirror base is provided with a rear plate portion extending to a cabin inner side. A gap is formed between the frame front side portion and the rear plate portion. An outer sealing plate portion overlapping with a cabin outer side of the mirror base is provided with a gap filling portion extending downward.

Abstract: A data generation device includes one or more processors. The processors input input data into a neural network and obtain an inference result of the neural network The processors calculate a first loss and a second loss. The first loss becomes smaller in value as a degree of matching between the inference result and a target label becomes larger. The target label indicates a correct answer of the inference. The second loss is a loss based on a contribution degree to the inference result of a plurality of elements included in the input data and the target label. The processors update the input data based on the first loss and the second loss.