Abstract: A processor-implemented method includes, based on scene information about a space, generating a first object set for objects associated with a first unit space among one or more unit spaces of the space, generating a first object affordance information set for the first object set by obtaining object affordance information for each of the objects of the first object set, generating first space affordance information corresponding to the first unit space by inputting the first object affordance information set for the first object set to a first model, and in response to receiving a task of a robot associated with the space, generating a movement path for accomplishing the task of the robot based on the first space affordance information.

Abstract: A method of controlling a companion robot includes receiving a multi-modal input signal from a user, extracting at least one feature vector from the multi-modal input signal using a universal encoder that maps information on different modalities to the same feature space, retrieving a knowledge database (DB) including a plurality of knowledge graphs based on the at least one feature vector, wherein the plurality of knowledge graphs includes at least one knowledge graph that accumulates experience data related to the user, inferring a causality vector for determining an action of the companion robot based on the at least one feature vector and a retrieval result of the knowledge DB, generating at least one sub-graph corresponding to the at least one knowledge graph based on the causality vector, and updating the at least one knowledge graph based on the generated at least one sub-graph.

Abstract: A method of determining a final path of a moving object includes acquiring pieces of sensor data, determining a first path of the moving object based on the pieces of sensor data, inputting at least one piece of sensor data among the pieces of sensor data into an encoder that encodes the at least one piece of sensor data, inputting the encoded at least one piece of sensor data into a generative neural network model that generates guide information on a path of the moving object, and determining the final path of the moving object, based on the first path and the guide information.

Abstract: A processor-implemented method with data processing using a neural network includes: determining a first translated image by translating a first image based on a second image, the first image and a second image that having different distortions, such that a distortion of the first image corresponds to a distortion of the second image; determining a first retranslated image by translating the first translated image such that a distortion of the first translated image corresponds to a distortion of the first image; and training a first deformation field generator configured to determine a first relative deformation field that represents a relative deformation from the first image to the second image and a second deformation field generator configured to determine a second relative deformation field that represents a relative deformation from the second image to the first image, based on a loss between the first retranslated image and the first image.

Abstract: A processor-implemented method with data processing using a neural network includes: determining a first translated image by translating a first image based on a second image, the first image and a second image that having different distortions, such that a distortion of the first image corresponds to a distortion of the second image; determining a first retranslated image by translating the first translated image such that a distortion of the first translated image corresponds to a distortion of the first image; and training a first deformation field generator configured to determine a first relative deformation field that represents a relative deformation from the first image to the second image and a second deformation field generator configured to determine a second relative deformation field that represents a relative deformation from the second image to the first image, based on a loss between the first retranslated image and the first image.

Abstract: A processor-implemented method with data processing using a neural network includes: determining a first translated image by translating a first image based on a second image, the first image and a second image that having different distortions, such that a distortion of the first image corresponds to a distortion of the second image; determining a first retranslated image by translating the first translated image such that a distortion of the first translated image corresponds to a distortion of the first image; and training a first deformation field generator configured to determine a first relative deformation field that represents a relative deformation from the first image to the second image and a second deformation field generator configured to determine a second relative deformation field that represents a relative deformation from the second image to the first image, based on a loss between the first retranslated image and the first image.