Abstract: An object tracking method includes inputting, into a neural network, two or more chronologically consecutive images, and matching similarity by comparing features extracted by the neural network, namely features of each of the two or more input images, and thereby outputting, as an identification result, identification information and position information about one or more objects depicted in a chronologically later image than a chronologically earlier image, which match one or more objects which are tracking candidates depicted in the chronologically earlier image. The neural network includes two or more identical structures having zero or more fully-connected layers and one or more convolution layers, and shares parameters among corresponding layers across the identical structures.

Abstract: Inputting an image to a neural network, performing convolution on a current frame included in the image to calculate a current feature map, which is a feature map at a present time, combining a past feature map, which is obtained by performing convolution on a past frame included in the image, and the current feature map, estimating an object candidate area using the combined past feature map and current feature map, estimating positional information and identification information regarding the one or more objects included in the current frame using the combined past feature map and current feature map and the estimated object candidate area, and outputting the positional information and the identification information regarding the one or more objects included in the current frame of the image estimated in the estimating as object detection results are included.

Abstract: An information processing method performed by a computer includes: obtaining a plurality of recognition result candidates in sensing data and a likelihood of each of the plurality of recognition result candidates, the plurality of recognition result candidates and the likelihood being obtained by inputting the sensing data to a model that is trained by machine learning and performs recognition processing; obtaining an indication designating a part to be analyzed in the sensing data; selecting at least one recognition result candidate from the plurality of recognition result candidates, based on (i) a relationship between each of the plurality of recognition result candidates and the part and (ii) the likelihood of each of the plurality of recognition result, candidates; and outputting the at least one recognition result candidate that is selected.

Abstract: An information processing method performed using a computer includes: obtaining first sensor data (SD) that is SD of a scene and includes noise; inputting the first SD to a single converter, and obtaining second SD outputted from the single converter as a result of denoising performed on the first SD by the single converter; obtaining third SD that is SD of a scene identical or corresponding to the scene, does not include the noise, and is different from the second SD; obtaining feature information of the second SD and feature information of the third SD, based on the second SD and the third SD, respectively; and training the single converter by machine learning using the second SD and the feature information of the second SD as converted data, and using the third SD and the feature information of the third SD as reference data corresponding to the converted data.

Abstract: An information processing method includes: obtaining noise region estimation information output from a first converter by a first image including a noise region being input to the first converter; obtaining a second image, on which noise region removal processing has been performed, output from a second converter by the noise region estimation information and the first image being input to the second converter; generating a fourth image including the estimated noise region by using the noise region estimation information and a third image including no noise region and a scene corresponding to the first image; training the first converter by using machine learning in which the first image is reference data and the fourth image is conversion data; and training the second converter by using machine learning in which the third image is reference data and the second image is conversion data.

Abstract: A determination method for determining the structure of a convolutional neural network includes acquiring N filters having the weights trained using a training image group as the initial values, where N is a natural number greater than or equal to 1, and increasing the number of the filters from N to M, where M is a natural number greater than or equal to 2 and is greater than N, by adding a filter obtained by performing a transformation used in image processing fields on at least one of the N filters.

Abstract: An information processing method includes: obtaining, from a mobile object, (i) at least one of operation information of the mobile object and traveling information indicating a traveling route taken by the mobile object, and (ii) sensor data obtained by sensing of the traveling route using a sensor included in the mobile object; causing a learning model to learn, using (i) the at least one of the operation information and the traveling information, and (ii) the sensor data as input, at least one of operation performed on and movement made by the mobile object using the sensor data obtained by sensing of the traveling route; generating, according to a number of times the learning model is trained, presentation information indicating the input required before the learning model completes learning about the traveling route; and causing a presentation device to present the presentation information.

Abstract: In an image recognition method executed by a computer of an image recognizer using a convolutional neural network, the convolutional neural network is a first convolutional neural network in which a fully connected layer is changed to a convolutional layer, and the method includes controlling a first convolutional neural network to acquire an input image, controlling the first convolutional neural network to estimate a center area of a recognition target in the acquired input image and to output a value indicating the estimated center area as a location of the recognition target in the input image.

Abstract: An information processing method includes: inputting an input tensor indicating data to a processor having a memory; causing the processor to perform, after elements of the input tensor are subjected to precomputation for conversion into a power-of-two format and are stored in the memory, convolution operation processing with only addition and shift operations by using the precomputed elements of the input tensor stored in the memory and weight tensors that are pre-converted into the power-of-two format in accordance with a predetermined algorithm, that are stored in the memory, and that indicate weights having a possibility of being used for a convolution operation; and outputting, as an output tensor, the elements of the input tensor on which the convolution operation processing is performed.

Abstract: An information processing apparatus includes an inputter, a comparison processor, and an outputter. The inputter inputs, in a neural network, a first data item that is one of data items included in time-series data. The comparison processor performs comparison between a first predicted data item predicted by the neural network and a second data item that is included in the time-series data. The first predicted data item is predicted as a data item first time after the first data item. The second data item is a data item the first time after the first data item. The outputter outputs information indicating warning if an error between the second data item and the first predicted data item is larger than a threshold after the comparison processor performs the comparison.

Abstract: A learning method includes an input process to input, to a neural network, a first image and a second image that constitute a moving image and that are temporally adjacent to each other, where the second image is an image subsequent to the first image with a predetermined time interval therebetween, a learning process to cause the neural network to use the first image and the second image and learn to output a transformation matrix applied to all pixels of the first image and used to convert the first image into the second image, and an output process to output, as a result of estimation of motion between the first image and the second image, a motion amount image generated from the transformation matrix and representing an amount of motion of each of the pixels of the first image that continues until the predetermined time interval elapses.

Abstract: Inputting an image to a neural network, performing convolution on a current frame included in the image to calculate a current feature map, which is a feature map at a present time, combining a past feature map, which is obtained by performing convolution on a past frame included in the image, and the current feature map, estimating an object candidate area using the combined past feature map and current feature map, estimating positional information and identification information regarding the one or more objects included in the current frame using the combined past feature map and current feature map and the estimated object candidate area, and outputting the positional information and the identification information regarding the one or more objects included in the current frame of the image estimated in the estimating as object detection results are included.

Abstract: In a data processing method executed by a computer: inputting, in a third trained model, first output data corresponding to first input data for a first trained model to obtain second output data, the third trained model being acquired through training in which (i) output data of the first trained model is used as training data, and (ii) output data of a second trained model acquired by converting the first trained model is used as label data; obtaining first label data of the first input data; and retraining the first trained model using first differential data corresponding to differences between the second output data and the first label data.

Abstract: An object tracking method includes inputting, into a neural network, two or more chronologically consecutive images, and matching similarity by comparing features extracted by the neural network, namely features of each of the two or more input images, and thereby outputting, as an identification result, identification information and position information about one or more objects depicted in a chronologically later image than a chronologically earlier image, which match one or more objects which are tracking candidates depicted in the chronologically earlier image. The neural network includes two or more identical structures having zero or more fully-connected layers and one or more convolution layers, and shares parameters among corresponding layers across the identical structures.

Abstract: A method, executed by a processor of an image generation system, includes obtaining an image of a first area included in a first image and an image of a second area included in a second image, calculating a first conversion parameter for converting the image of the first area such that color information regarding the image of the first area becomes similar to color information regarding the image of the second area, converting the first image using the first conversion parameter, and generating a third image as a training image used for machine learning for image recognition by combining the converted first image and the second image with each other.

Abstract: Provided is an image tagging device including: a first functional unit in which an image including an object that is a target of privacy protection is stored and that removes privacy information by changing part of the master image; a second functional unit that acquires the changed image from the first functional unit and changes a region image of an object that is not to be tagged in the first image; a function that distributes the changed image from the second functional unit to a tagging operation terminal device and receives image tag information from the tagging operation terminal device over a network; and a tagged image generator that generates a tagged image on the basis of the master image and the image tag information. This makes it possible to collect tagged images while achieving both privacy protection and an improvement in efficiency of a tagging operation.

Abstract: An image recognition method includes: receiving an image; acquiring processing result information including values of processing results of convolution processing at positions of a plurality of pixels that constitute the image by performing the convolution processing on the image by using different convolution filters; determining 1 feature for each of the positions of the plurality of pixels on the basis of the values of the processing results of the convolution processing at the positions of the plurality of pixels included in the processing result information and outputting the determined feature for each of the positions of the plurality of pixels; performing recognition processing on the basis of the determined feature for each of the positions of the plurality of pixels; and outputting recognition processing result information obtained by performing the recognition processing.

Abstract: An image transmitted, through a network, from any of at least one terminal having a function of capturing an image or obtaining an image from another device is obtained. A probability that the obtained image includes a certain imaging target is calculated. If the probability is higher than a first threshold, information indicating the certain imaging target is added to the image. If the probability is lower than a second threshold, the information indicating the certain imaging target is not to the image. If the probability is equal to or higher than the second threshold and if the probability is equal to or lower than the first threshold, the image and request reception information for requesting addition of the information is transmitted to the image to any of the at least one terminal through the network.

Abstract: Learning method includes performing a first process in which a coarse class classifier configured with a first neural network is made to classify a plurality of images given as a set of images each attached with a label indicating a detailed class into a plurality of coarse classes including a plurality of detailed classes and is then made to learn a first feature that is a feature common in each of the coarse classes, and performing a second process in which a detailed class classifier, configured with a second neural network that is the same in terms of layers other than the final layer as but different in terms of the final layer from the first neural network made to perform the learning in the first process, is made to classify the set of images into detailed classes and learn a second feature of each detailed class.

Abstract: A determination method for determining the structure of a convolutional neural network includes acquiring N filters having the weights trained using a training image group as the initial values, where N is a natural number greater than or equal to 1, and increasing the number of the filters from N to M, where M is a natural number greater than or equal to 2 and is greater than N, by adding a filter obtained by performing a transformation used in image processing fields on at least one of the N filters.