Abstract: An image processing system includes one or more memories; and one or more processors coupled to the one or more memories and the one or more processors configured to acquire a degree of influence of each region of image data on a recognition result when recognition processing is performed for the image data, and reduce data size of the image data based on the degree of influence.

Abstract: A method includes: generating a refine image from an incorrect image from which an incorrect label is inferred by a neural network; generating a third map by superimposing a first map and a second map, the first map indicating pixels to each of which a change is made in generating the refine image, of plural pixels in the incorrect image, the second map indicating a degree of attention for each local region in the refine image, each local region being a region that has drawn attention at the time of inference by the neural network, and the third map indicating a degree of importance for each pixel for inferring a correct label; and obtaining an added value for respective divided region in the third map by summing pixel values within the respective divided region, the respective divided region being a region divided according to a predetermined index.

Abstract: An analysis method executed by a computer includes acquiring a refine image that maximizes a score for inferring a correct label by an inferring process using a trained model, the refine image being generated from an input image used when an incorrect label is inferred; generating a map indicating a region of pixels having the same or similar level of attention degree related to inference in the inferring process, of a plurality of pixels in the generated refine image, based on a feature amount used in the inferring process; extracting an image corresponding to a pixel region whose level in the generated map is a predetermined level, from calculated images calculated based on the input image and the refine image; and generating an output image that specifies a portion related to an inference error in the inferring process, among the calculated images, based on image processing on the extracted image.

Abstract: An information processing apparatus includes: a memory; and a processor coupled to the memory and configured to: partition pixel values in a unit of row of an input image into a plurality of sections and allocates threads to the respective sections of the row, the threads being enabled to run in parallel by a processor; calculate, with each of the threads allocated in each row, distances each from a pixel having a certain value in the corresponding section of the row in the input image, and generates a first distance image which stores values indicating the distances; and calculate, with each of the threads allocated in each row, a first boundary value indicating a distance from a pixel having the certain value in another section of each row, by using a calculation result of the first boundary value in the another section of each row.

Abstract: An image processing apparatus, includes a memory; and a processor coupled to the memory and configured to: generate a trained machine learning model by learning a machine learning model using a first set of image data, output an inference result by inputting a second set of image data to the trained machine learning model, and process a region of interest at a time of inference with respect to image data for which an inference result is correct in the second set of image data.

Abstract: A method includes: generating a refine image from an incorrect image from which an incorrect label is inferred by a neural network; generating a third map by superimposing a first map and a second map, the first map indicating pixels to each of which a change is made in generating the refine image, of plural pixels in the incorrect image, the second map indicating a degree of attention for each local region in the refine image, each local region being a region that has drawn attention at the time of inference by the neural network, and the third map indicating a degree of importance for each pixel for inferring a correct label; and obtaining an added value for respective divided region in the third map by summing pixel values within the respective divided region, the respective divided region being a region divided according to a predetermined index.

Abstract: A method includes: generating a refine image having a maximized correct label score of inference from an incorrect image from which an incorrect label is inferred by a neural network; generating a third map by superimposing a first map and a second map, the first map indicating pixels to each of which a change is made in generating the refine image, of a plurality of pixels of the incorrect image, the second map indicating a degree of attention for each local region in the refine image, the each local region being a region that has drawn attention by the neural network; and specifying a set of pixels that cause incorrect inference in the incorrect image by calculating a pixel value of the third map for each set of pixels, wherein the map generating processing adjusts the second map based on appearance frequency of each degree of attention.

Abstract: A method includes: generating a refine image having a maximized correct label score of inference from an incorrect image by which an incorrect label is inferred by a neural network; generating a third map by superimposing a first map and a second map, the first map indicating pixels to each of which a change is made in generating the refine image, of plural pixels of the incorrect image, the second map indicating a degree of attention for each local region in the refine image, the each local region being a region that has drawn attention at the time of inference by the neural network, and the third map indicating a degree of importance for each pixel for inferring a correct label; and specifying an image section based on a pixel value of the third map, the image section corresponding to a region causing incorrect inference in the incorrect image.

Abstract: An analysis method executed by a computer includes acquiring a refine image that maximizes a score for inferring a correct label by an inferring process using a trained model, the refine image being generated from an input image used when an incorrect label is inferred; generating a map indicating a region of pixels having the same or similar level of attention degree related to inference in the inferring process, of a plurality of pixels in the generated refine image, based on a feature amount used in the inferring process; extracting an image corresponding to a pixel region whose level in the generated map is a predetermined level, from calculated images calculated based on the input image and the refine image; and generating an output image that specifies a portion related to an inference error in the inferring process, among the calculated images, based on image processing on the extracted image.

Abstract: An analysis method implemented by a computer includes: generating a refine image by changing an incorrect inference image such that a correct label score of inference is maximized, the incorrect inference image being an input image when an incorrect label is inferred in an image recognition process; and narrowing, based on a score of a label, a predetermined region to specify an image section that causes incorrect inference, the score of the label being inferred by inputting to an inferring process an image obtained by replacing the predetermined region in the incorrect inference image with the refine image.

Abstract: An information processing apparatus includes a memory and a processor which changes a frame rate in a period of moving picture data to a first set value of the frame rate, codes respective pictures in the period to obtain a code amount according to the first set value, calculates a cumulative value of a code amount of each kind of pictures different in applicable coding processing regarding the respective pictures included in the period, calculates a ratio of the cumulative value, estimates a change tendency of a difficulty level of coding of respective pictures in a next period after the period based on transition of the frame rate and transition of the ratio of the cumulative value, and obtains a second set value of the frame rate and a third set value of a bit rate applied to the next period according to the change tendency of the difficulty level.

Abstract: An apparatus reconstructs a three dimensional shape of an object from a plurality of images captured from different viewpoints by visual hull. The apparatus stores, into a first memory provided for the apparatus, silhouette image data of the object extracted from the plurality of images, and determines a group of a plurality of epipolar lines having inclinations within a predetermined range in images designated as reference views among the plurality of images, based on positions of neighboring silhouette pixels in an image designated as a target view among the plurality of images. The apparatus further determines, based on the determined group and a capacity of a cache provided for the apparatus, a cache area to be used in a search for a line segment indicating a range where the object exists in each of the plurality of epipolar lines in the group.

Abstract: An information processing apparatus includes a processor. the processor determines a sign of a first value on a basis of a result of an operation of a sum of products of a plurality of color-difference residuals and a plurality of luminance residuals, and calculates the first value on a basis of a first product of a sum of luminance residuals that are negative values among the plurality of luminance residuals and a constant when the sign of the first value is determined to be positive, and calculates the first value on a basis of a second product of a sum of luminance residuals that are positive values among the plurality of luminance residuals and the constant when the sign of the first value is determined to be negative.

Abstract: An information processing apparatus includes: a memory; and a processor coupled to the memory and configured to: partition pixel values in a unit of row of an input image into a plurality of sections and allocates threads to the respective sections of the row, the threads being enabled to run in parallel by a processor; calculate, with each of the threads allocated in each row, distances each from a pixel having a certain value in the corresponding section of the row in the input image, and generates a first distance image which stores values indicating the distances; and calculate, with each of the threads allocated in each row, a first boundary value indicating a distance from a pixel having the certain value in another section of each row, by using a calculation result of the first boundary value in the another section of each row.

Abstract: A moving image encoding apparatus includes one or more processors configured to calculate encoding costs regarding each of a plurality of blocks having different block sizes on the basis of differences between original images of the plurality of blocks and specific prediction images of the plurality of blocks respectively wherein the plurality of blocks is defined by recursively dividing an encoding target image included in a moving image, perform a selection of a first block size in a first hierarchy among the plurality of hierarchies by comparing the encoding costs with each other, perform a selection of a first intra-prediction mode for a first block having the first block size, and encode the first block on the basis of a difference between a first prediction image of the first block generated by the first intra-prediction mode and a first original image of the first block.

Abstract: A method for surface-finishing one or more metal liners is provided, the method comprising chemical milling to remove wrinkled textures generated during the plastic deformation of the metal liners.

Abstract: An apparatus sequentially acquires, from a plurality of reference imaging devices for imaging a silhouette imaged with a base imaging device from viewpoints different from a viewpoint of the base imaging device, silhouette existing position information based on the reference imaging devices, and transforms the silhouette existing position information into a common coordinate system, where the silhouette existing position information indicates an existing position of the silhouette. The apparatus detects a silhouette absence range in which the silhouette does not exist, based on a result of comparison of the silhouette existing position information acquired this time and the silhouette existing position information acquired last time, and searches a range in which the silhouette exists, based on the silhouette absence range.

Abstract: An apparatus obtains a second straight-line by mapping a first straight-line passing a projection center of a target-image taken by a first imaging-device in a three-dimensional space and a point representing an object in a projection-plane of the first imaging-device, onto each of reference-images respectively taken by second imaging-devices, and generates a line-segment representing an existing-range of the object on the second straight-line for each reference-image. The apparatus stores the coordinate-value of the second-endpoint and a difference between coordinate-values of the first-endpoint and a second-endpoint of the line-segment in a memory.

Abstract: An apparatus reconstructs a three dimensional shape of an object from a plurality of images captured from different viewpoints by visual hull. The apparatus stores, into a first memory provided for the apparatus, silhouette image data of the object extracted from the plurality of images, and determines a group of a plurality of epipolar lines having inclinations within a predetermined range in images designated as reference views among the plurality of images, based on positions of neighboring silhouette pixels in an image designated as a target view among the plurality of images. The apparatus further determines, based on the determined group and a capacity of a cache provided for the apparatus, a cache area to be used in a search for a line segment indicating a range where the object exists in each of the plurality of epipolar lines in the group.

Abstract: An information processing apparatus includes a memory and a processor which changes a frame rate in a period of moving picture data to a first set value of the frame rate, codes respective pictures in the period to obtain a code amount according to the first set value, calculates a cumulative value of a code amount of each kind of pictures different in applicable coding processing regarding the respective pictures included in the period, calculates a ratio of the cumulative value, estimates a change tendency of a difficulty level of coding of respective pictures in a next period after the period based on transition of the frame rate and transition of the ratio of the cumulative value, and obtains a second set value of the frame rate and a third set value of a bit rate applied to the next period according to the change tendency of the difficulty level.