Abstract: According to one embodiment, an image analysis device includes one or more processors configured to receive input of an image; calculate feature amount information indicating a feature of a region of the image; recognize a known object from the image on the basis of the feature amount information, the known object being registered in learning data of image recognition; recognize a generalization object from the image on the basis of the feature amount information, the generalization object being generalizable from the known object; and output output information on an object identified from the image as the known object or the generalization object.

Abstract: According to one embodiment, an electronic device includes one or more processors. The one or more processors obtain an image captured by a camera with a filter having a first area transmitting light of a first wavelength range and a second area transmitting light of a second wavelength range. The image includes a first color-component image based on the light of the first wavelength range and a second color-component image based on the light of the second wavelength range. The one or more processors notify a user of an effective area for calculation of depth information based on a bias of color information in the first color-component image and the second color-component image.

Abstract: According to one embodiment, a learning method for causing a second statistical model to learn using a first statistical model is provided. The method includes obtaining a first learning image, cutting out each local area of the obtained first learning image, and obtaining a first prediction value output from the first statistical model by inputting each local area to the first statistical model and obtaining a second prediction value output from the second statistical model by inputting the entire area of the first learning image to the second statistical model, and causing the second statistical model to learn based on a difference between the first prediction value and the second prediction value.

Abstract: According to one embodiment, a robot motion planning device includes processing circuitry. The processing circuitry receives observation information obtained by observing at least part of a movable range of a robot. The processing circuitry determines, in a case where first observation information is received, a target position to which the robot is to make a motion, using an action-value function and the first observation information. The processing circuitry receives measurement information obtained by measuring a state of the robot, calculates a difference corresponding to the first observation information, using the measurement information, and determines a motion plan of a force-controlled motion of the robot, based on the target position and the difference.

Abstract: According to one embodiment, an image processing device includes a hardware processor implemented by one or more processors. The hardware processor is configured to acquire an image captured by a monocular camera. The hardware processor is configured to estimate a distance to an object based on the acquired image. The hardware processor is configured to calculate the distance to the object by considering an atmosphere failure influencing the acquired image and executing a statistical process on the estimated distance.

Abstract: According to one embodiment, an image processing device includes storage and a processor. The storage stores a statistical model generated by learning bokeh produced in a first image affected by aberration of an optical system, the bokeh changing nonlinearly in accordance with a distance to a subject in the first image. The processor obtains a second image affected by the aberration of the optical system. The processor inputs the second image to the statistical model and obtains distance information indicating a distance to a subject in the second image.

Abstract: According to one embodiment, a learning method for causing a second statistical model to learn using a first statistical model is provided. The method includes obtaining a first learning image, cutting out each local area of the obtained first learning image, and obtaining a first prediction value output from the first statistical model by inputting each local area to the first statistical model and obtaining a second prediction value output from the second statistical model by inputting the entire area of the first learning image to the second statistical model, and causing the second statistical model to learn based on a difference between the first prediction value and the second prediction value.

Abstract: According to one embodiment, an optical filter associated with a lens unit having a first end includes a filter part, a filter frame, and an attachment part. The filter part includes a first filter area that transmits light rays of a first wavelength band and a second filter area that transmits light rays of a second wavelength band. The first wavelength band and the second wavelength band include a common wavelength band. The filter frame holds the filter part. The attachment part is coupled with the filter frame and is detachably attached to the first end of the lens unit. When the attachment part is attached to the first end of the lens unit, a distance between the filter part and an aperture within the lens unit is longer than a distance between the first end and the aperture.

Abstract: According to one embodiment, an image processing device includes storage and a processor. The storage stores a statistical model generated by learning bokeh produced in a first image affected by aberration of an optical system, the bokeh changing nonlinearly in accordance with a distance to a subject in the first image. The processor obtains a second image affected by the aberration of the optical system. The processor inputs the second image to the statistical model and obtains distance information indicating a distance to a subject in the second image.

Abstract: According to one embodiment, a robot motion planning device includes processing circuitry. The processing circuitry receives observation information obtained by observing at least part of a movable range of a robot. The processing circuitry determines, in a case where first observation information is received, a target position to which the robot is to make a motion, using an action-value function and the first observation information. The processing circuitry receives measurement information obtained by measuring a state of the robot, calculates a difference corresponding to the first observation information, using the measurement information, and determines a motion plan of a force-controlled motion of the robot, based on the target position and the difference.

Abstract: An information processing device according to one embodiment includes a first receiver, a second receiver, a first converter, a second converter, and a calculator. The first receiver receives input of first data belonging to a first modality. The second receiver receives input of second data belonging to a second modality that is different from the first modality. The first converter converts the first data into a first representation representing a point or a first area in a D-dimensional vector space (D is a natural number). The second converter converts the second data into a second representation representing a second area in the D-dimensional vector space. The calculator calculates similarity between the first data and the second data by using the first representation and the second representation.

Abstract: According to one embodiment, an electronic device includes one or more processors. The one or more processors obtain an image captured by a camera with a filter having a first area transmitting light of a first wavelength range and a second area transmitting light of a second wavelength range. The image includes a first color-component image based on the light of the first wavelength range and a second color-component image based on the light of the second wavelength range. The one or more processors notify a user of an effective area for calculation of depth information based on a bias of color information in the first color-component image and the second color-component image.

Abstract: According to one embodiment, an optical filter associated with a lens unit having a first end includes a filter part, a filter frame, and an attachment part. The filter part includes a first filter area that transmits light rays of a first wavelength band and a second filter area that transmits light rays of a second wavelength band. The first wavelength band and the second wavelength band include a common wavelength band. The filter frame holds the filter part. The attachment part is coupled with the filter frame and is detachably attached to the first end of the lens unit. When the attachment part is attached to the first end of the lens unit, a distance between the filter part and an aperture within the lens unit is longer than a distance between the first end and the aperture.

Abstract: According to one embodiment, an image processing device includes a hardware processor implemented by one or more processors, The hardware processor is configured to acquire an image captured by a monocular camera. The hardware processor is configured to estimate a distance to an object based on the acquired image. The hardware processor is configured to calculate the distance to the object by considering an atmosphere failure influencing the acquired image and executing a statistical process on the estimated distance.

Abstract: According to one embodiment, an image analysis device includes one or more processors configured to receive input of an image; calculate feature amount information indicating a feature of a region of the image; recognize a known object from the image on the basis of the feature amount information, the known object being registered in learning data of image recognition; recognize a generalization object from the image on the basis of the feature amount information, the generalization object being generalizable from the known object; and output output information on an object identified from the image as the known object or the generalization object.

Abstract: According to an embodiment, an information processing device includes a memory and processing circuitry. The processing circuitry is configured to receive an image and area information indicating a first area. The processing circuitry is configured to detect an object from the image. The processing circuitry is configured to, based on positional relationship between the first area and the object, estimate probability of occurrence of an event in the first area.

Abstract: According to an embodiment, an image output device includes an extractor, a search unit, an associate unit, and a controller. The extractor is configured to extract a first parameter that varies in accordance with a movement of an object from at least one first image of the object, and extract a second parameter that varies in accordance with a movement of the object from each second image of the object. The search unit is configured to search for a second parameter similar to the first parameter. The associate unit is configured to associate the first image from which the first parameter is extracted with the second image from which the second parameter that is retrieved with respect to the first parameter is extracted. The controller is configured to instruct an output unit to output an image based on the first and second images that are associated to each other.

Abstract: According to an embodiment, an information processing device includes a memory and processing circuitry. The processing circuitry is configured to receive an image and area information indicating a first area. The processing circuitry is configured to detect an object from the image. The processing circuitry is configured to, based on positional relationship between the first area and the object, estimate probability of occurrence of an event in the first area.

Abstract: According to an embodiment, a device includes a memory and processing circuitry. When time taken by a user to carry out first behavior is equal to or more than a first threshold, or when number of times the first behavior is repeated is equal to or more than a second threshold, the processing circuitry is configured to output first guidance to lead the user to first expected behavior that is behavior the user is expected to carry out subsequent to the first behavior. When the time taken to carry out the first behavior is less than the first threshold, or when the number of times the first behavior is repeated is less than the second threshold, the processing circuitry is configured to omit the first guidance to lead the user to the first expected behavior or outputs second guidance that is simpler than the first guidance.

Abstract: An image processing method includes: retaining marker information including markers, each of the markers having a type which is shape, pattern or color; acquiring images in which at least one of the markers is caught; referring the marker information to detect the type and a position of the marker caught in the image; and dividing the image into a plurality of divided areas on the basis of the positions of one or more markers in the image, the plurality of divided areas having no common area and each including at least one type of the marker.