Abstract: A gravity compensation assembly including a main frame having one side connected to a weight body to be supported, and the other side where an inner rotation portion and an outer rotation portion spaced apart from each other and having rotation axis in an x-axis direction are coupled, an auxiliary frame having one side rotatably connected to the inner rotation portion of the main frame, and the other side where an auxiliary rotation portion having a rotation axis in the x-axis direction is coupled, and an elastic force providing having one side rotatably connected to the outer rotation portion of the main frame, and the other side coupled to the auxiliary rotation portion of the auxiliary frame, and configured to perform load compensation by an elastic force when a center of gravity changes as a relative angle of the main frame and the auxiliary frame changes may be provided.

Abstract: A system for detecting and describing keypoints in images is described. A camera is configured to capture an image including a plurality of pixels. A fully convolutional network is configured to jointly and concurrently: generate descriptors for each of the pixels, respectively; generate reliability scores for each of the pixels, respectively; and generate repeatability scores for each of the pixels, respectively. A scoring module is configured to generate scores for the pixels, respectively, based on the reliability scores and the repeatability scores of the pixels, respectively. A keypoint list module is configured to: select X of the pixels having the X highest scores, where X is an integer greater than 1; and generate a keypoint list including: locations of the selected X pixels; and the descriptors of the selected X pixels.

Abstract: A method for position estimation of a vehicle based on a graph structure according to an exemplary embodiment of the present invention may include a sensing step of creating sensing data according to translation of the vehicle using a plurality of sensors; a graph structure creating step of creating a node according to a reference period and setting a constraint condition between the nodes to the sensing data synchronized with the reference period to create a graph structure; a graph structure modifying step of modifying the graph structure using the asynchronous sensing data when the asynchronous sensing data which is not synchronized with the reference period among the sensing data is input; and a position estimating step of creating position information of the vehicle using the graph structure.

Abstract: A pose estimation training system includes: a first model configured to generate a first 6 degrees of freedom (DoF) pose of a first camera that captured a first image from a first domain; a second model configured to generate a second 6 DoF pose of a second camera that captured a second image from a second domain, where the second domain is different than the first domain; a discriminator module configured to, based on first and second outputs from the first and second encoder modules, generate a discriminator output indicative of whether the first and second images are from the same domain; and a training control module configured to, based on the discriminator output, selectively adjust at least one weight value shared by the first model and the second model.

Abstract: An end device including a communication module configured to wirelessly communicate with an edge server managed by a cloud server, and at least one processor connected to the communication module, the at least one processor configured to receive a control command from the edge server through the communication module, and operate according to the control command.

Abstract: Disclosed are an electronic device and an operating method of the electronic device, which relate to location-based augmented reality (AR) linkage of object-based augmentation content, and may recognize an object based on an image being captured, detect a preset location in association with at least one of the object and the image, determine augmentation content based on the object and the location, and output the augmentation content in correspondence to the object while displaying the image.

Abstract: Electronic devices and/or operating methods of the electronic device to provide visual localization based on outdoor three-dimensional (3D) map information may be provided. Such electronic devices may be configured to acquire two-dimensional (2D) image information about an outdoor environment, generate 3D map information about the outdoor environment based on the 2D image information, and determine a position of a point in the 3D map information corresponding to a query image.

Abstract: Disclosed is a robot hand comprising a base, a first bracket provided on an upper portion of the base, a second bracket provided on an upper portion of the base and configured such that the other side of the second bracket is rotatable in a horizontal direction about one side of the second bracket, a plurality of finger modules coupled to upper portions of the first and second brackets, respectively, and configured to tilt in order to grip an object, finger drive units installed on the first and second brackets and configured to transmit driving power to the finger modules and a rotation drive unit installed at one side of the second bracket and configured to transmit driving power so that the other side of the second bracket is rotatable in the horizontal direction about one side of the second bracket.

Abstract: Disclosed is a robot joint device comprising first and second plates positioned in parallel with each other, links each having a first end connected to the first plate and a second end connected to the second plate, connecting members configured to connect the two ends of each of the links and the first and second plates, respectively, so that angles and rotations of the links are adjustable relative to the first and second plates, a rotary shaft having two ends penetrating the first and second plates and rotatably installed, a gear reduction unit installed in the first plate and connected to the first end of the rotary shaft, a pulley connected to the second end of the rotary shaft and configured to transmit driving power to the rotary shaft, and a drive unit configured to transmit the driving power to the pulley.

Abstract: A system for detecting and describing keypoints in images is described. A camera is configured to capture an image including a plurality of pixels. A fully convolutional network is configured to jointly and concurrently: generate descriptors for each of the pixels, respectively; generate reliability scores for each of the pixels, respectively; and generate repeatability scores for each of the pixels, respectively. A scoring module is configured to generate scores for the pixels, respectively, based on the reliability scores and the repeatability scores of the pixels, respectively. A keypoint list module is configured to: select X of the pixels having the X highest scores, where X is an integer greater than 1; and generate a keypoint list including: locations of the selected X pixels; and the descriptors of the selected X pixels.

Abstract: A training system for training a trained model for use by a navigating robot to perform visual navigation includes memory including N base virtual training environments, each of the N base virtual training environments including a field of view at a location within an indoor space, where N is an integer greater than 1. A randomization module is configured to generate N varied virtual training environments based on the N base virtual training environments, respectively, by varying at least one characteristic of the respective N base virtual training environments. A training module is configured to train the trained model for use by the navigating robot to perform visual navigation based on a training set including: the N base virtual training environments; and the N varied virtual training environments.

Abstract: A method of identifying a current location, performed by a server having wireless communication with a terminal, includes receiving from the terminal, surrounding environment information input in natural language by a user; performing natural language processing on the surrounding environment information to extract place information about a reference place input by the user and positional relationship information about a positional relationship between the reference place and the user; searching a map database for the reference place corresponding to the place information; estimating a candidate region in which the user is likely to be located, based on the searched-for reference place and the positional relationship information; and determining the candidate region as the user's current location. Because the current location is identified based on the surrounding environment information input by the user, the user's current location can be identified more accurately.

Abstract: A method and system for controlling an indoor autonomous robot. A method of controlling an indoor autonomous driving of a robot in a cloud system includes receiving, by the cloud system, first sensing data that is generated by a mapping robot autonomously driving in a target facility using a first sensor that detects the first sensing data with respect to an inside of the target facility; generating an indoor map of the target facility based on the first sensing data; receiving, from a service robot present in the target facility through a network, second sensing data that is generated by the service robot using a second sensor that detects the second sensing data with respect to the inside of the target facility; and controlling an indoor autonomous driving of the service robot with respect to the target facility based on the second sensing data and the generated indoor map.

Abstract: Provided are methods and systems for automatically collecting and updating information that is related to a point of interest in real space. According to the methods, information relating to a plurality of points of interest (POIs), which exist in real space, is automatically collected and compared to previously collected information, and if there is a change, the change can be automatically updated so as to provide a location-based service, such as a map service, in real space such as a downtown street or a mall.

Abstract: A method of training a predictor to predict a location of a computing device in an indoor environment incudes: receiving training data including strength of signals received from wireless access points at positions of an indoor environment, where the training data includes: a subset of labeled data including signal strength values and location labels; and a subset of unlabeled data including signal strength values and not including labels indicative of locations; training a variational autoencoder to minimize a reconstruction loss of the signal strength values of the training data, where the variational autoencoder includes encoder neural networks and decoder neural networks; and training a classification neural network to minimize a prediction loss on the labeled data, where the classification neural network generates a predicted location based on the latent variable, and where the encoder neural networks and the classification neural network form the predictor.

Abstract: An electrically driven moving vehicle includes a main body on which an article to be transported is loaded, and a handle provided on one side of the main body. A pair of force sensor units are provided in the handle. A force applied by a user is sensed by the pair of force sensor units to control the operation of electric motors. The user can operate the electrically driven moving vehicle while holding a position convenient for the user rather than a specific position of the handle. In addition, the user can conveniently operate the electrically driven moving vehicle with only one hand, without using both hands.