Abstract: A data generation method for generating data for domain generalization in machine learning includes performing, with a computer, augmentation using training data as raw data usable to train a machine learning model, and extracting, with the computer, a dataset including the raw data and data generated through the augmentation as a dataset for the domain generalization.

Abstract: The observation update unit (152) increases a weight of a particle for a particle representing a position and a posture closer to a position and a posture of an operation object indicated by an observation result, and increases a weight of a corresponding particle for a particle closer to a state in which an object of interest and the operation object arranged in a position and a posture represented by each of particles are in contact with each other in a virtual space where shapes and a relative positional relationship of the object of interest and the operation object are expressed in a case in which the observation result indicates that the contact has occurred. A state estimation unit (154) calculates an estimated state that is a position and a posture of the operation object estimated based on a set of the particles in which a weight of each of the particles is adjusted.

Abstract: An inference apparatus provides target data to multiple inference models to cause the inference models each derived from local learning data obtained in a different environment to perform predetermined inference to obtain an inference result from each of the inference models. The inference apparatus determines the value of each combining parameter using environment data, weights the inference result from each of the inference models using the determined value of each combining parameter, and combines the weighted inference result from each inference model together to generate an inference result in a target environment.

Abstract: Provided is an image determination device. The image determination device is provided with: feature extractors which output, on the basis of an image to be examined, each piece of feature data indicating a specific feature of the image; a determiner which outputs, on the basis of the feature data output from the extractors, output data indicating the determination result pertaining to the image; and a training part which trains the determiner so as to output, output data indicating the label data associated with the training image on the basis of the feature data output when the training image is input to the extractors, wherein the training part further trains, by using new training data, the determiner so that the output data indicating the label data associated with the image is output by the determiner.

Abstract: An image determination device includes feature extractors which output, on the basis of an image to be examined, each piece of feature data indicating a specific feature of the image; a first training part which trains a first determiner so as to output first output data indicating first label data associated with a first training image on the basis of first feature data output when the first training image is input to the feature extractors; a second training part which trains a second determiner so as to output second output data indicating second label data associated with a second training image on the basis of second feature data output when the second training image is input to the feature extractors; and an output part which outputs, on the basis of the first output data and the second output data, output data indicating an overall determination result about an image.

Abstract: An image determination device includes: a training model which outputs, on the basis of an image to be examined, output data indicating a determination result about the image; a training part which trains the training model to output, by using training data including a training image and label data, output data indicating the label data associated with the training image, when the training image is input to the training model; a dividing part which divides the training data into a plurality of pieces of sub-training data; a measurement part which measures accuracy of determination when the training part trains the training model by using each of the plurality of pieces of sub-training data; and selection part which selects at least any one among the plurality of pieces of sub-training data on the basis of the accuracy of determination.

Abstract: An image determination device includes: a training model which outputs, on the basis of an image to be examined, output data indicating a determination result about the image; a training part which trains the training model to output, by using training data including a training image and label data, output data indicating the label data associated with the training image, when the training image is input to the training model; a dividing part which divides the training data into a plurality of pieces of sub-training data; a measurement part which measures accuracy of determination when the training part trains the training model by using each of the plurality of pieces of sub-training data; and selection part which selects at least any one among the plurality of pieces of sub-training data on the basis of the accuracy of determination.

Abstract: A joint structure includes a locking mechanism for switching a free state in which a second element is independent from a first element and capable of moving, and a locked state in which the second element is fixed to the first element. The locking mechanism includes a first member joined to the first element, a second member joined to the second element, and a flexible wire-shaped member in which one end thereof is attached to the second member and another end thereof is led out to the outside of the joint structure via a through hole provided in the first member. The joint structure enters the locked state by the wire-shaped member being pulled to bring the second member into contact with the first member, and enters the free state by the wire-shaped member being fed to separate the second member from the first member.

Abstract: A prediction model (1) includes a first module (M1) that calculates, for each of a plurality of objects (xi) in a dataset (x), an index value (vi) corresponding to a combination of the object (xi) and attribute information (a) using a neural network, and a second module (M2) that calculates a prediction result (y) of an operation to be performed by a user by performing a predetermined process on a plurality of index values (v1, . . . , vN) obtained from the first module (M1) and corresponding to the respective plurality of objects (x1, . . . , xN).

Abstract: Provided is a robot system comprising: a robot mechanism provided with fingers for grasping a workpiece; a detection unit that detects a grasping state of the workpiece by the fingers; a recognition unit that recognizes the number of workpieces being grasped by the fingers on the basis of a detection result from the detection unit; and a control unit that, after causing the robot mechanism to perform a motion of grasping the workpieces from a placing area where a plurality of workpieces are placed, if the number of the workpieces recognized by the recognition unit is different from a specified number specified in advance, controls the motion of the robot mechanism such that the number of the grasped workpieces is equal to the specified number. At least one of the fingers is flexible. The control unit controls the motion of the robot mechanism such that, in a state in which workpieces are being grasped by the fingers, the position of at least one of the fingers is changed.

Abstract: A control device sets a relative relation amount between a plurality of target objects that are to be final objectives, repeatedly acquires observation data from a sensor and calculates the relative relation amount between the plurality of target objects existing in the environment from the acquired observation data. Further, the control device determines a series of the relative relation amounts in a state that is to be an objective from the relative relation amount at a time point at which control of the behavior starts until the relative relation amount as the final objectives is realized, and repeatedly determines control instructions so as to change the relative relation amount in a present state calculated from the latest observation data into the relative relation amount in a state of an objective to be transitioned to next. Then, the control device outputs the determined control instruction to a robot device.

Abstract: A control apparatus of a robot may include a state obtaining unit configured to obtain state observation data including flexible related observation data, which is observation data regarding a state of at least one of a flexible portion, a portion of the robot on a side where an object is gripped relative to the flexible portion, and the gripped object; and a controller configured to control the robot so as to output an action to be performed by the robot to perform predetermined work on the object, in response to receiving the state observation data, based on output obtained as a result of inputting the state observation data obtained by the state obtaining unit to a learning model, the learning model being learned in advance through machine learning and included in the controller.

Abstract: Provided is a technique capable of generating a discriminator having high inference precision at low cost. An inspection system according to one aspect of the present invention generates, from first image data, plural pieces of second image data of a product; determines, by a first discriminator, whether to adopt each piece of the second image data as learning data; generates a learning data set by the second image data that is determined to be adopted as learning data; constructs, by applying machine learning using the generated learning data set, a second discriminator that determines the acceptability of a product; and uses the constructed second discriminator to determine the acceptability of a product in target image data.

Abstract: A joint structure includes a locking mechanism for switching a free state in which a second element is independent from a first element and capable of moving, and a locked state in which the second element is fixed to the first element. The locking mechanism includes a first member joined to the first element, a second member joined to the second element, and a flexible wire-shaped member in which one end thereof is attached to the second member and another end thereof is led out to the outside of the joint structure via a through hole provided in the first member. The joint structure enters the locked state by the wire-shaped member being pulled to bring the second member into contact with the first member, and enters the free state by the wire-shaped member being fed to separate the second member from the first member.

Abstract: A control apparatus of a robot that includes a flexible portion and a locking mechanism for fixing the flexible portion, the control apparatus including: a lock control unit configured to control locking and unlocking of the flexible portion; and an operation control unit configured to control operation of the robot using different types of control policies depending on whether the flexible portion is locked or unlocked. With this control apparatus, it is possible to effectively control a robot in which locking and unlocking of the flexible portion can be switched.

Abstract: A technique capable of implementing more efficient connection between users is provided. A matching system includes: storage means configured to hold a question set including a plurality of questions; acquisition means configured to present one or more questions included in the question set to a user and acquires one or more answers to the presented questions from the user; first matching processing means configured to execute first matching processing based on one or more questions answered by the user; and second matching processing means configured to execute second matching processing based on content of one or more answers included in the one or more answers from the user when a result of the first matching processing satisfies a first determination condition.

Abstract: A robot control apparatus according to one or more embodiments may include: a calculating unit configured to calculate an interference range of a robot based on a model of the robot in a state in which an object is gripped by a gripper with which the robot is equipped; and a planning unit configured to plan a motion of the robot based on the model and the interference range.

Abstract: An inference apparatus provides target data to multiple inference models to cause the inference models each derived from local learning data obtained in a different environment to perform predetermined inference to obtain an inference result from each of the inference models. The inference apparatus determines the value of each combining parameter using environment data, weights the inference result from each of the inference models using the determined value of each combining parameter, and combines the weighted inference result from each inference model together to generate an inference result in a target environment.

Abstract: The present disclosure relates to a method, an apparatus and a system for controlling a robot, and a storage medium. The method uses a neural network connected to an external memory to conduct the controlling of the robot, and comprises: inputting input data into the learned neural network to obtain output data, wherein said input data comprises an image about an object, said output data comprises control data about said robot; and establishing an association between part or all of the information generated by said neural network during the calculation and said input data and/or said output data, wherein said part or all of the information represents a feature of said object related to said control data. Thus, the user can grasp the calculation process of the neural network.

Abstract: This neural network-type image processing device is provided with an input layer which comprises one unit where an input image is inputted, an output layer which comprises one unit where an output image is outputted, and multiple intermediate layers which are arranged between the input layer and the output layer and each of which comprises multiple units, the unit of the input layer, the units of the intermediate layers, and the unit of the output layer are fully connected with connection coefficients. The units of the intermediate layers are image processing modules which perform image processing on the image inputted to said units. The input image is inputted from the unit of the input layer, passes through the units of the intermediate layers, and is then outputted as an output image from the unit of the output layer; the connection coefficients are updated with learning based on a backpropagation algorithm.