Patent
Copyright © 2026 Matter42. All rights reserved.
Patent
Atlas literature
Patent
Patent drawings and their descriptions. Click a drawing to enlarge it.
Claims define the patent's legal scope. Independent claims stand alone; dependent claims (nested) narrow them. Click a claim to expand its dependents.
Alearning method for learning reduction of distortion occurred in a warped image generated in a process of stabilizing ajittered image by using a GAN (Generative Adversarial Network) including a generating network and a discriminating network, comprising steps of (a) a lea rn ing device, if at least one initial image is acquired, instructing an adjusting layer included in the generating network to adjust at least part of initial feature values corresponding to pixels included in the initial image, to thereby transform the initial image into at least one adjusted image; and (b) the learning device, if at least part of(i) at least one naturality score representing at least one probability of the adjusted image being determined as real by the discriminating network, (ii) at least one maintenance score representing whether one or more initial objects included in the initial image are included in the adjusted image with their characteristics maintained, and (iii) at least one similarity score representing a degree of similarity between the initial image and the adjusted image are acquired, instructing a loss layer included in the generating network to generate a generating network loss by referring to said at least part of the naturality score, the maintenance score and the similarity score, and lea rn parameters of the generating network by backpropagating the generating network loss.
The lea rn ing method of Claim 1, wherein, at the step of(b), the discriminating network determines said at least one probability of the adjusted image being real or fake by referring to a feature map corresponding to the adjusted image, to thereby generate the naturality score.
The lea rn ing method of Claim 1, wherein, at the step of(b), an object detection network generates one or more class scores on one or more ROIs corresponding to one or more adjusted objects included in the adjusted image, and generate the maintenance score by referring to the class scores.
The lea rn ing method of Claim 1, wherein, at the step of(b), a comparing layer included in the generating network generates the similarity score by referring to information on differences between the initial feature values and their corresponding adjusted feature values included in the adjusted image.
The lea rn ing method of Claim 1, wherein, at the step of(b), the generating network loss allows the parameters included in the generating network to be learned to make an integrated score calculated by referring to at least part of the naturality score, the maintenance score and the similarity score be larger.
A testing method for testing reduction of distortion occurred in a warped image generated in a process of stabilizing a jittered image by using a GAN(Generative Adversarial Network) including a generating network and a discriminating network, comprising a step of on condition that (1) a lea rn ing device has instructed an adjusting layer included in the generating network to adjust at least part of initial feature values for training corresponding to pixels included in at least one initial training image, to thereby transform the initial training image into at least one adjusted training image, and (2) the learning device has instructed a loss layer included in the generating network to generate a generating network loss by referring to at least part of (i) at least one naturality score representing at least one probability of the adjusted training image being determined as real by the discriminating network, (ii) at least one maintenance score representing whether one or more initial objects included in the initial training image are included in the adjusted training image with their characteristics maintained, and (iii) at least one similarity score representing a degree of similarity between the initial training image and the adjusted training image, and learn parameters of the generating network by backpropagating the generating network loss; a testing device instructing the adjusting layer included in the generating network to adjust at least part of initial feature values for testing corresponding to pixels included in at least one initial test image, to thereby transform the initial test image into at least one adjusted test image.
Alea m ing device for learning reduction of distortion occurred in a warped image generated in a process of stabilizing ajittered image by using a GAN(Generative Adversarial Network) including a generating network and a discriminating network, comprising: at least one memory that stores instructions; and at least one processor configured to execute the instructions to: perform processes of (I) if at least one initial image is acquired, instructing an adjusting layer included in the generating network to adjust at least part of initial feature values corresponding to pixels included in the initial image, to thereby transform the initial image into at least one adjusted image, and (I I) if at least part of(i) at least one naturality score representing at least one probability of the adjusted image being determined as real by the discriminating network, (ii) at least one maintenance score representing whether one or more initial objects included in the initial image are included in the adjusted image with their characteristics maintaine d, and (iii) at least one similarity score representing a degree of similarity between the initial image and the adjusted image are acquire d, instructing a loss layer included in the generating network to generate a generating network loss by referring to said at least part of the naturality score, the maintenance score and the similarity score, and lea rn parameters of the generating network by backpropagating the generating network loss.
The learning device of Claim 10, wherein, at the process of(II), the discriminating network determines said at least one probability of the adjusted image being real or fake by referring to a feature map corresponding to the adjusted image, to thereby generate the naturality score.
The learning device of Claim 10, wherein, at the process of (II), an object detection network generates one or more class scores on one or more ROIs corresponding to one or more adjusted objects included in the adjusted image, and generate the maintenance score by referring to the class scores.
The learning device of Claim 10, wherein, at the process of (II), a comparing layer included in the generating network generates the similarity score by referring to information on differences between the initial feature values and their corresponding adjusted feature values included in the adjusted image.
The learning device of Claim 10, wherein, at the process of (II), the generating network loss allows the parameters included in the generating network to be learned to make an integrated score calculated by referring to at least part of the naturality score, the maintenance score and the similarity score be larger.
Atesting device for reduction of distortion occurred in a warped image generated in a process of stabilizing a jittered image by using a GAN(Generative Adversarial Network) including a generating network and a discriminating network, comprising: at least one memory that stores instructions; and at least one processor, on condition that (1) a lea rn ing device has instructed an adjusting layer included in the generating network to adjust at least part of initial feature values for training corresponding to pixels included in at least one initial training image, to thereby transform the initial training image into at least one adjusted training image, and (2) the learning device has instructed a loss layer included in the generating network to generate a generating network loss by referring to at least part of(i) at least one naturality score representing at least one probability of the adjusted training image being determined as real by the discriminating network, (ii) at least one maintenance score representing whether one or more initial objects included in the initial training image are included in the adjusted training image with their characteristics maintained, and (iii) at least one similarity score representing a degree of similarity between the initial training image and the adjusted training image, and lea rn parameters of the generating network by backpropagating the generating network loss; configured to execute the instructions to: perform a process of instructing the adjusting layer included in the generating network to adjust at least part of initial feature values for testing corresponding to pixels included in at least one initial test image, to thereby transform the initial test image into at least one adjusted test image.
Patent
Atlas literature
Patent
Patent drawings and their descriptions. Click a drawing to enlarge it.
Claims define the patent's legal scope. Independent claims stand alone; dependent claims (nested) narrow them. Click a claim to expand its dependents.
Alearning method for learning reduction of distortion occurred in a warped image generated in a process of stabilizing ajittered image by using a GAN (Generative Adversarial Network) including a generating network and a discriminating network, comprising steps of (a) a lea rn ing device, if at least one initial image is acquired, instructing an adjusting layer included in the generating network to adjust at least part of initial feature values corresponding to pixels included in the initial image, to thereby transform the initial image into at least one adjusted image; and (b) the learning device, if at least part of(i) at least one naturality score representing at least one probability of the adjusted image being determined as real by the discriminating network, (ii) at least one maintenance score representing whether one or more initial objects included in the initial image are included in the adjusted image with their characteristics maintained, and (iii) at least one similarity score representing a degree of similarity between the initial image and the adjusted image are acquired, instructing a loss layer included in the generating network to generate a generating network loss by referring to said at least part of the naturality score, the maintenance score and the similarity score, and lea rn parameters of the generating network by backpropagating the generating network loss.
The lea rn ing method of Claim 1, wherein, at the step of(b), the discriminating network determines said at least one probability of the adjusted image being real or fake by referring to a feature map corresponding to the adjusted image, to thereby generate the naturality score.
The lea rn ing method of Claim 1, wherein, at the step of(b), an object detection network generates one or more class scores on one or more ROIs corresponding to one or more adjusted objects included in the adjusted image, and generate the maintenance score by referring to the class scores.
The lea rn ing method of Claim 1, wherein, at the step of(b), a comparing layer included in the generating network generates the similarity score by referring to information on differences between the initial feature values and their corresponding adjusted feature values included in the adjusted image.
The lea rn ing method of Claim 1, wherein, at the step of(b), the generating network loss allows the parameters included in the generating network to be learned to make an integrated score calculated by referring to at least part of the naturality score, the maintenance score and the similarity score be larger.
A testing method for testing reduction of distortion occurred in a warped image generated in a process of stabilizing a jittered image by using a GAN(Generative Adversarial Network) including a generating network and a discriminating network, comprising a step of on condition that (1) a lea rn ing device has instructed an adjusting layer included in the generating network to adjust at least part of initial feature values for training corresponding to pixels included in at least one initial training image, to thereby transform the initial training image into at least one adjusted training image, and (2) the learning device has instructed a loss layer included in the generating network to generate a generating network loss by referring to at least part of (i) at least one naturality score representing at least one probability of the adjusted training image being determined as real by the discriminating network, (ii) at least one maintenance score representing whether one or more initial objects included in the initial training image are included in the adjusted training image with their characteristics maintained, and (iii) at least one similarity score representing a degree of similarity between the initial training image and the adjusted training image, and learn parameters of the generating network by backpropagating the generating network loss; a testing device instructing the adjusting layer included in the generating network to adjust at least part of initial feature values for testing corresponding to pixels included in at least one initial test image, to thereby transform the initial test image into at least one adjusted test image.
Alea m ing device for learning reduction of distortion occurred in a warped image generated in a process of stabilizing ajittered image by using a GAN(Generative Adversarial Network) including a generating network and a discriminating network, comprising: at least one memory that stores instructions; and at least one processor configured to execute the instructions to: perform processes of (I) if at least one initial image is acquired, instructing an adjusting layer included in the generating network to adjust at least part of initial feature values corresponding to pixels included in the initial image, to thereby transform the initial image into at least one adjusted image, and (I I) if at least part of(i) at least one naturality score representing at least one probability of the adjusted image being determined as real by the discriminating network, (ii) at least one maintenance score representing whether one or more initial objects included in the initial image are included in the adjusted image with their characteristics maintaine d, and (iii) at least one similarity score representing a degree of similarity between the initial image and the adjusted image are acquire d, instructing a loss layer included in the generating network to generate a generating network loss by referring to said at least part of the naturality score, the maintenance score and the similarity score, and lea rn parameters of the generating network by backpropagating the generating network loss.
The learning device of Claim 10, wherein, at the process of(II), the discriminating network determines said at least one probability of the adjusted image being real or fake by referring to a feature map corresponding to the adjusted image, to thereby generate the naturality score.
The learning device of Claim 10, wherein, at the process of (II), an object detection network generates one or more class scores on one or more ROIs corresponding to one or more adjusted objects included in the adjusted image, and generate the maintenance score by referring to the class scores.
The learning device of Claim 10, wherein, at the process of (II), a comparing layer included in the generating network generates the similarity score by referring to information on differences between the initial feature values and their corresponding adjusted feature values included in the adjusted image.
The learning device of Claim 10, wherein, at the process of (II), the generating network loss allows the parameters included in the generating network to be learned to make an integrated score calculated by referring to at least part of the naturality score, the maintenance score and the similarity score be larger.
Atesting device for reduction of distortion occurred in a warped image generated in a process of stabilizing a jittered image by using a GAN(Generative Adversarial Network) including a generating network and a discriminating network, comprising: at least one memory that stores instructions; and at least one processor, on condition that (1) a lea rn ing device has instructed an adjusting layer included in the generating network to adjust at least part of initial feature values for training corresponding to pixels included in at least one initial training image, to thereby transform the initial training image into at least one adjusted training image, and (2) the learning device has instructed a loss layer included in the generating network to generate a generating network loss by referring to at least part of(i) at least one naturality score representing at least one probability of the adjusted training image being determined as real by the discriminating network, (ii) at least one maintenance score representing whether one or more initial objects included in the initial training image are included in the adjusted training image with their characteristics maintained, and (iii) at least one similarity score representing a degree of similarity between the initial training image and the adjusted training image, and lea rn parameters of the generating network by backpropagating the generating network loss; configured to execute the instructions to: perform a process of instructing the adjusting layer included in the generating network to adjust at least part of initial feature values for testing corresponding to pixels included in at least one initial test image, to thereby transform the initial test image into at least one adjusted test image.
Patent
Atlas literature
Patent
Patent drawings and their descriptions. Click a drawing to enlarge it.
Claims define the patent's legal scope. Independent claims stand alone; dependent claims (nested) narrow them. Click a claim to expand its dependents.
Alearning method for learning reduction of distortion occurred in a warped image generated in a process of stabilizing ajittered image by using a GAN (Generative Adversarial Network) including a generating network and a discriminating network, comprising steps of (a) a lea rn ing device, if at least one initial image is acquired, instructing an adjusting layer included in the generating network to adjust at least part of initial feature values corresponding to pixels included in the initial image, to thereby transform the initial image into at least one adjusted image; and (b) the learning device, if at least part of(i) at least one naturality score representing at least one probability of the adjusted image being determined as real by the discriminating network, (ii) at least one maintenance score representing whether one or more initial objects included in the initial image are included in the adjusted image with their characteristics maintained, and (iii) at least one similarity score representing a degree of similarity between the initial image and the adjusted image are acquired, instructing a loss layer included in the generating network to generate a generating network loss by referring to said at least part of the naturality score, the maintenance score and the similarity score, and lea rn parameters of the generating network by backpropagating the generating network loss.
The lea rn ing method of Claim 1, wherein, at the step of(b), the discriminating network determines said at least one probability of the adjusted image being real or fake by referring to a feature map corresponding to the adjusted image, to thereby generate the naturality score.
The lea rn ing method of Claim 1, wherein, at the step of(b), an object detection network generates one or more class scores on one or more ROIs corresponding to one or more adjusted objects included in the adjusted image, and generate the maintenance score by referring to the class scores.
The lea rn ing method of Claim 1, wherein, at the step of(b), a comparing layer included in the generating network generates the similarity score by referring to information on differences between the initial feature values and their corresponding adjusted feature values included in the adjusted image.
The lea rn ing method of Claim 1, wherein, at the step of(b), the generating network loss allows the parameters included in the generating network to be learned to make an integrated score calculated by referring to at least part of the naturality score, the maintenance score and the similarity score be larger.
A testing method for testing reduction of distortion occurred in a warped image generated in a process of stabilizing a jittered image by using a GAN(Generative Adversarial Network) including a generating network and a discriminating network, comprising a step of on condition that (1) a lea rn ing device has instructed an adjusting layer included in the generating network to adjust at least part of initial feature values for training corresponding to pixels included in at least one initial training image, to thereby transform the initial training image into at least one adjusted training image, and (2) the learning device has instructed a loss layer included in the generating network to generate a generating network loss by referring to at least part of (i) at least one naturality score representing at least one probability of the adjusted training image being determined as real by the discriminating network, (ii) at least one maintenance score representing whether one or more initial objects included in the initial training image are included in the adjusted training image with their characteristics maintained, and (iii) at least one similarity score representing a degree of similarity between the initial training image and the adjusted training image, and learn parameters of the generating network by backpropagating the generating network loss; a testing device instructing the adjusting layer included in the generating network to adjust at least part of initial feature values for testing corresponding to pixels included in at least one initial test image, to thereby transform the initial test image into at least one adjusted test image.
Alea m ing device for learning reduction of distortion occurred in a warped image generated in a process of stabilizing ajittered image by using a GAN(Generative Adversarial Network) including a generating network and a discriminating network, comprising: at least one memory that stores instructions; and at least one processor configured to execute the instructions to: perform processes of (I) if at least one initial image is acquired, instructing an adjusting layer included in the generating network to adjust at least part of initial feature values corresponding to pixels included in the initial image, to thereby transform the initial image into at least one adjusted image, and (I I) if at least part of(i) at least one naturality score representing at least one probability of the adjusted image being determined as real by the discriminating network, (ii) at least one maintenance score representing whether one or more initial objects included in the initial image are included in the adjusted image with their characteristics maintaine d, and (iii) at least one similarity score representing a degree of similarity between the initial image and the adjusted image are acquire d, instructing a loss layer included in the generating network to generate a generating network loss by referring to said at least part of the naturality score, the maintenance score and the similarity score, and lea rn parameters of the generating network by backpropagating the generating network loss.
The learning device of Claim 10, wherein, at the process of(II), the discriminating network determines said at least one probability of the adjusted image being real or fake by referring to a feature map corresponding to the adjusted image, to thereby generate the naturality score.
The learning device of Claim 10, wherein, at the process of (II), an object detection network generates one or more class scores on one or more ROIs corresponding to one or more adjusted objects included in the adjusted image, and generate the maintenance score by referring to the class scores.
The learning device of Claim 10, wherein, at the process of (II), a comparing layer included in the generating network generates the similarity score by referring to information on differences between the initial feature values and their corresponding adjusted feature values included in the adjusted image.
The learning device of Claim 10, wherein, at the process of (II), the generating network loss allows the parameters included in the generating network to be learned to make an integrated score calculated by referring to at least part of the naturality score, the maintenance score and the similarity score be larger.
Atesting device for reduction of distortion occurred in a warped image generated in a process of stabilizing a jittered image by using a GAN(Generative Adversarial Network) including a generating network and a discriminating network, comprising: at least one memory that stores instructions; and at least one processor, on condition that (1) a lea rn ing device has instructed an adjusting layer included in the generating network to adjust at least part of initial feature values for training corresponding to pixels included in at least one initial training image, to thereby transform the initial training image into at least one adjusted training image, and (2) the learning device has instructed a loss layer included in the generating network to generate a generating network loss by referring to at least part of(i) at least one naturality score representing at least one probability of the adjusted training image being determined as real by the discriminating network, (ii) at least one maintenance score representing whether one or more initial objects included in the initial training image are included in the adjusted training image with their characteristics maintained, and (iii) at least one similarity score representing a degree of similarity between the initial training image and the adjusted training image, and lea rn parameters of the generating network by backpropagating the generating network loss; configured to execute the instructions to: perform a process of instructing the adjusting layer included in the generating network to adjust at least part of initial feature values for testing corresponding to pixels included in at least one initial test image, to thereby transform the initial test image into at least one adjusted test image.
Patent
Atlas literature
Patent
Patent drawings and their descriptions. Click a drawing to enlarge it.
Claims define the patent's legal scope. Independent claims stand alone; dependent claims (nested) narrow them. Click a claim to expand its dependents.
Alearning method for learning reduction of distortion occurred in a warped image generated in a process of stabilizing ajittered image by using a GAN (Generative Adversarial Network) including a generating network and a discriminating network, comprising steps of (a) a lea rn ing device, if at least one initial image is acquired, instructing an adjusting layer included in the generating network to adjust at least part of initial feature values corresponding to pixels included in the initial image, to thereby transform the initial image into at least one adjusted image; and (b) the learning device, if at least part of(i) at least one naturality score representing at least one probability of the adjusted image being determined as real by the discriminating network, (ii) at least one maintenance score representing whether one or more initial objects included in the initial image are included in the adjusted image with their characteristics maintained, and (iii) at least one similarity score representing a degree of similarity between the initial image and the adjusted image are acquired, instructing a loss layer included in the generating network to generate a generating network loss by referring to said at least part of the naturality score, the maintenance score and the similarity score, and lea rn parameters of the generating network by backpropagating the generating network loss.
The lea rn ing method of Claim 1, wherein, at the step of(b), the discriminating network determines said at least one probability of the adjusted image being real or fake by referring to a feature map corresponding to the adjusted image, to thereby generate the naturality score.
The lea rn ing method of Claim 1, wherein, at the step of(b), an object detection network generates one or more class scores on one or more ROIs corresponding to one or more adjusted objects included in the adjusted image, and generate the maintenance score by referring to the class scores.
The lea rn ing method of Claim 1, wherein, at the step of(b), a comparing layer included in the generating network generates the similarity score by referring to information on differences between the initial feature values and their corresponding adjusted feature values included in the adjusted image.
The lea rn ing method of Claim 1, wherein, at the step of(b), the generating network loss allows the parameters included in the generating network to be learned to make an integrated score calculated by referring to at least part of the naturality score, the maintenance score and the similarity score be larger.
A testing method for testing reduction of distortion occurred in a warped image generated in a process of stabilizing a jittered image by using a GAN(Generative Adversarial Network) including a generating network and a discriminating network, comprising a step of on condition that (1) a lea rn ing device has instructed an adjusting layer included in the generating network to adjust at least part of initial feature values for training corresponding to pixels included in at least one initial training image, to thereby transform the initial training image into at least one adjusted training image, and (2) the learning device has instructed a loss layer included in the generating network to generate a generating network loss by referring to at least part of (i) at least one naturality score representing at least one probability of the adjusted training image being determined as real by the discriminating network, (ii) at least one maintenance score representing whether one or more initial objects included in the initial training image are included in the adjusted training image with their characteristics maintained, and (iii) at least one similarity score representing a degree of similarity between the initial training image and the adjusted training image, and learn parameters of the generating network by backpropagating the generating network loss; a testing device instructing the adjusting layer included in the generating network to adjust at least part of initial feature values for testing corresponding to pixels included in at least one initial test image, to thereby transform the initial test image into at least one adjusted test image.
Alea m ing device for learning reduction of distortion occurred in a warped image generated in a process of stabilizing ajittered image by using a GAN(Generative Adversarial Network) including a generating network and a discriminating network, comprising: at least one memory that stores instructions; and at least one processor configured to execute the instructions to: perform processes of (I) if at least one initial image is acquired, instructing an adjusting layer included in the generating network to adjust at least part of initial feature values corresponding to pixels included in the initial image, to thereby transform the initial image into at least one adjusted image, and (I I) if at least part of(i) at least one naturality score representing at least one probability of the adjusted image being determined as real by the discriminating network, (ii) at least one maintenance score representing whether one or more initial objects included in the initial image are included in the adjusted image with their characteristics maintaine d, and (iii) at least one similarity score representing a degree of similarity between the initial image and the adjusted image are acquire d, instructing a loss layer included in the generating network to generate a generating network loss by referring to said at least part of the naturality score, the maintenance score and the similarity score, and lea rn parameters of the generating network by backpropagating the generating network loss.
The learning device of Claim 10, wherein, at the process of(II), the discriminating network determines said at least one probability of the adjusted image being real or fake by referring to a feature map corresponding to the adjusted image, to thereby generate the naturality score.
The learning device of Claim 10, wherein, at the process of (II), an object detection network generates one or more class scores on one or more ROIs corresponding to one or more adjusted objects included in the adjusted image, and generate the maintenance score by referring to the class scores.
The learning device of Claim 10, wherein, at the process of (II), a comparing layer included in the generating network generates the similarity score by referring to information on differences between the initial feature values and their corresponding adjusted feature values included in the adjusted image.
The learning device of Claim 10, wherein, at the process of (II), the generating network loss allows the parameters included in the generating network to be learned to make an integrated score calculated by referring to at least part of the naturality score, the maintenance score and the similarity score be larger.
Atesting device for reduction of distortion occurred in a warped image generated in a process of stabilizing a jittered image by using a GAN(Generative Adversarial Network) including a generating network and a discriminating network, comprising: at least one memory that stores instructions; and at least one processor, on condition that (1) a lea rn ing device has instructed an adjusting layer included in the generating network to adjust at least part of initial feature values for training corresponding to pixels included in at least one initial training image, to thereby transform the initial training image into at least one adjusted training image, and (2) the learning device has instructed a loss layer included in the generating network to generate a generating network loss by referring to at least part of(i) at least one naturality score representing at least one probability of the adjusted training image being determined as real by the discriminating network, (ii) at least one maintenance score representing whether one or more initial objects included in the initial training image are included in the adjusted training image with their characteristics maintained, and (iii) at least one similarity score representing a degree of similarity between the initial training image and the adjusted training image, and lea rn parameters of the generating network by backpropagating the generating network loss; configured to execute the instructions to: perform a process of instructing the adjusting layer included in the generating network to adjust at least part of initial feature values for testing corresponding to pixels included in at least one initial test image, to thereby transform the initial test image into at least one adjusted test image.