Fujifilm Corporation

Yoshiro Kitamura

Wataru Ito

Keigo Nakamura

Sadato Akahori

Akimichi Ichinose

Kensuke Terakawa

Akira Kudo

Yasunori Ohta

Naoyuki Nishino

Eiichi Kito

Yuanzhong Li

Daiki Harada

Naoki Mochizuki

Deepak Keshwani

Akio Sato

Tsutomu Inoue

Mizuki Takei

Saeko Sasuga

Yuichi Kurihara

An image generation device derives, for a subject including a specific structure, a subject model representing the subject by deriving each feature amount of the target image having the at least one representation format and combining the feature amounts based on the target image. A latent variable derivation unit derives a latent variable obtained by dimensionally compressing a feature of the subject model according to the target information based on the target information and the subject model. A virtual image derivation unit outputs a virtual image having the representation format represented by the target information based on the target information, the subject model, and the latent variable.

Learning device, learning method, learning program, image generation device, image generation method, image generation program, and image generation model

Provided are a learning method and a learning system of a generative model, a program, a learned model, and a super resolution image generating device that can handle input data of any size and can suppress the amount of calculation at the time of image generation. A learning method according to an embodiment of the present disclosure is a learning method for performing machine learning of a generative model that estimates, from a first image, a second image including higher resolution image information than the first image, the method comprising using a generative adversarial network including a generator which is the generative model and a discriminator which is an identification model that identifies whether provided data is data of a correct image for learning or data derived from an output from the generator and implementing a self-attention mechanism only in a network of the discriminator among the generator and the discriminator.

Learning method, learning system, learned model, program, and super resolution image generating device

Provided is a machine learning device and method that enables machine learning of labeling, in which a plurality of labels are attached to volume data at one effort with excellent accuracy, using training data having label attachment mixed therein. A probability calculation unit () calculates a value (soft label) indicating a likelihood of labeling of a class Ci for each voxel of a second slice image by means of a learned teacher model (). A detection unit () detects 'bronchus' and “blood vessel” for the voxels of the second slice image using a known method, such as a region expansion method and performs labeling of “bronchus” and “blood vessel”. A correction probability setting unit () replaces the soft label with a hard label of “bronchus” or “blood vessel” detected by the detection unit (). A distillation unit () performs distillation of a student model () from the teacher model () using the soft label after correction by means of the correction probability setting unit (). With this, the learned student model () is obtained.

Machine learning device and method

A region specification apparatus specifies a region of an object which is included in an input image and which includes a plurality of subclass objects having different properties. The region specification apparatus includes a first discriminator that specifies an object candidate included in the input image. The first discriminator has a component configured to predict at least one of movement or transformation of a plurality of anchors according to the property of the subclass object and specify an object candidate region surrounding the object candidate.

Region specification apparatus, region specification method, region specification program, learning apparatus, learning method, learning program, and discriminator

Provided are an image processing apparatus, an image processing method, and a program that can suppress an error in the segmentation of a medical image. An image processing apparatus includes: a segmentation unit () that applies deep learning to perform segmentation which classifies a medical image () into a specific class on the basis of a local feature of the medical image; and a global feature classification unit () that applies deep learning to classify the medical image into a global feature which is an overall feature of the medical image. The segmentation unit shares a weight of a first low-order layer which is a low-order layer with a second low-order layer which is a low-order layer in the global feature classification unit.

Image processing apparatus, image processing method, and program

Provided are a linear structure extraction device, a method, a program, and a learned model which can detect a linear structure in an image. A linear structure extraction device according to an embodiment of the present disclosure includes a learning model that is learned to receive an input of the image and output, as a prediction result, one or more element points which constitute the linear structure from the image, in which the learning model includes a first processing module that receives the image and generates a feature map representing a feature amount of the image by convolution processing, and a second processing module that calculates a shift amount from a unit center point to the element point of the linear structure closest to the unit center point, for each unit obtained by dividing the feature map into a plurality of the units including regions having a predetermined size in a grid pattern.

Linear structure extraction device, method, program, and learned model

Provided are a learning method, a learning device, a generative model, and a program that generate an image including high resolution information without adjusting a parameter and largely correcting a network architecture even in a case in which there is a variation of the parts of an image to be input. Only a first image is input to a generator of a generative adversarial network that generates a virtual second image having a relatively high resolution by using the first image having a relatively low resolution, and a second image for learning or the virtual second image and part information of the second image for learning or the virtual second image are input to a discriminator that identifies the second image for learning and the virtual second image.

Learning method, learning device, generative model, and program

A machine learning method and an apparatus, a program, a learned model, and a discrimination apparatus capable of controlling a calculation amount by learning a new task without changing output performance for an existing task in a learned network are provided. A machine learning method according to one aspect of the present disclosure includes a step of adding a new feature amount to at least one intermediate layer included in a learned first neural network that has learned a task of performing first class classification, a step of generating a second neural network having a structure in which a network structure of a calculation path of an existing feature amount of the first neural network is maintained and a new feature amount of a subsequent layer is calculated by performing processing of convolving each of the existing feature amount and the new feature amount, and a step of causing the second neural network to acquire a processing function of performing second class classification by performing learning of the second neural network using a set of learning data corresponding to the second class classification.

Machine learning method and apparatus, program, learned model, and discrimination apparatus using multilayer neural network

A learning support deviceincludes an acquisition unit, a registration unit, a storage device, a learning unit, and a controller. The acquisition unitacquires an image of a region of interest and a name of the region of interest by analyzing an interpretation report. The registration unitregisters training data consisting of the image of the region of interest and the name of the region of interest acquired by the acquisition unitin the storage device. The learning unitperforms learning for generating a discrimination model, which outputs the image of the region of interest and the name of the region of interest with respect to an input of an inspection imageof the interpretation report, using a plurality of pieces of training dataregistered in the storage device

Learning support device, learning support method, learning support program, region-of-interest discrimination device, region-of-interest discrimination method, region-of-interest discrimination program, and learned model

A prediction apparatus includes a learning section that performs machine learning in which, with respect to a combination of different types of captured images obtained by imaging the same subject, one captured image is set to an input and another captured image is set to an output to generate a prediction model; and a controller that performs a control for inputting a first image to the prediction model as an input captured image and outputting a predicted second image that is a captured image having a type different from that of the input captured image.

Prediction apparatus, prediction method, prediction program

A determination processing unit determines a disease region in a medical image including an axisymmetric structure. A first determination section of the determination processing unit generates a feature amount map of the medical image from the medical image. A second determination section second inverts the feature amount map with reference to a symmetry axis to generate an inverted feature amount map. A third determination section superimposes the feature amount map and the inverted feature amount map on each other and determines the disease region in the medical image using the feature amount map and the inverted feature amount map superimposed on each other.

Medical image processing apparatus, medical image processing method, and medical image processing program

An extraction unit extracts a target object from a three-dimensional image, and a feature point detection unit detects at least one feature point included in the three-dimensional image. A reference axis setting unit sets a reference axis in the three-dimensional image based on the feature amount, and a two-dimensional image generation unit generates a two-dimensional image by projecting a target object, which is included in the three-dimensional image, in a specific projection direction using the reference axis as a reference. A classification unit classifies each pixel of the target object into a plurality of classes based on the two-dimensional image.

Image classification apparatus, method, and program

[Objective] To enable a three dimensional image to be accurately classified into a plurality of classes with a small amount of calculations, in an image classifying apparatus, an image classifying method, and an image classifying program. [Constitution] A three dimensional image is classified into a plurality of classes by a convolutional neural network, in which a plurality of processing layers are connected hierarchically. The convolutional neural network includes: a convoluting layer that administers a convoluting process on each of a plurality of two dimensional images, which are generated by the neural network administering a projection process on the three dimensional image according to a plurality of processing parameters; and a pooling layer that pools the values of the same position within each of the plurality of two dimensional images which have undergone the convoluting process.

Image classifying apparatus, image classifying method, and image classifying program

For assigning a binary label representing belonging to a target region or not to each pixel in an image: a predicted shape of the target region is set; a pixel group including N pixels is selected, where N is a natural number of 4 or more, which have a positional relationship representing the predicted shape; and an energy function is set, which includes an N-th order term in which a variable is a label of each pixel of the pixel group, so that a value of the N-th order term is at a minimum value when a combination of the labels assigned to the pixels of the pixel group is a pattern matching the predicted shape, and increases in stages along with an increase in a number of pixels to which a label different from the pattern is assigned. The labeling is performed by minimizing the energy function.

Image processing apparatus, and operation method and program therefor

When binary labeling is performed, an outline specification unit specifies a first outline present toward a target region and a second outline present toward a non-target region, and which have shapes similar to an outline of the target region. A voxel selection unit selects an N number of voxels constituting all of the first outline and the second outline. The energy setting unit sets N-order energy when a condition that all of the voxels of the first outline belong to the target region and all of the voxels of the second outline belong to the non-target region is satisfied smaller than the N-order energy when the condition is not satisfied. After then, labeling is performed by minimizing energy.

Image processing apparatus, non-transitory computer-readable recording medium having stored therein image processing program, and operation method of image processing apparatus

Growing community of inventors

Tokyo, Japan

Yoshiro Kitamura