The method of describing deformation camouflage spots based on feature space has some shortcomings,such as inaccurate description and difficult reproduction.Depending on the strong fitting ability of the generative ad...The method of describing deformation camouflage spots based on feature space has some shortcomings,such as inaccurate description and difficult reproduction.Depending on the strong fitting ability of the generative adversarial network model,the distribution of deformation camouflage spot pattern can be directly fitted,thus simplifying the process of spot extraction and reproduction.The requirements of background spot extraction are analyzed theoretically.The calculation formula of limiting the range of image spot pixels is given and two kinds of spot data sets,forestland and snowfield,are established.Spot feature is decomposed into shape,size and color features,and a GAN(Generative Adversarial Network)framework is established.The effects of different loss functions on network training results are analyzed in the experiment.In the meantime,when the input dimension of generator network is 128,the balance between sample diversity and quality can be achieved.The effects of sample generation are investigated in two aspects.Subjectively,the probability of the generated spots being distinguished in the background is counted,and the results are all less than 20% and mostly close to zero.Objectively,the features of the spot shape are calculated and the independent sample T-test is applied to verify that the features are from the same distribution,and all the P-Values are much higher than 0.05.Both subjective and objective methods prove that the spots generated by this method are similar to the background spots.The proposed method can directly generate the desired camouflage pattern spots,which provides a new technical method for the deformation camouflage pattern design and camouflage effect evaluation.展开更多
It is important to understand how ballistic materials respond to impact from projectiles such that informed decisions can be made in the design process of protective armour systems. Ballistic testing is a standards-ba...It is important to understand how ballistic materials respond to impact from projectiles such that informed decisions can be made in the design process of protective armour systems. Ballistic testing is a standards-based process where materials are tested to determine whether they meet protection, safety and performance criteria. For the V50ballistic test, projectiles are fired at different velocities to determine a key design parameter known as the ballistic limit velocity(BLV), the velocity above which projectiles perforate the target. These tests, however, are destructive by nature and as such there can be considerable associated costs, especially when studying complex armour materials and systems. This study proposes a unique solution to the problem using a recent class of machine learning system known as the Generative Adversarial Network(GAN). The GAN can be used to generate new ballistic samples as opposed to performing additional destructive experiments. A GAN network architecture is tested and trained on three different ballistic data sets, and their performance is compared. The trained networks were able to successfully produce ballistic curves with an overall RMSE of between 10 and 20 % and predicted the V50BLV in each case with an error of less than 5 %. The results demonstrate that it is possible to train generative networks on a limited number of ballistic samples and use the trained network to generate many new samples representative of the data that it was trained on. The paper spotlights the benefits that generative networks can bring to ballistic applications and provides an alternative to expensive testing during the early stages of the design process.展开更多
Ground military target recognition plays a crucial role in unmanned equipment and grasping the battlefield dynamics for military applications, but is disturbed by low-resolution and noisyrepresentation. In this paper,...Ground military target recognition plays a crucial role in unmanned equipment and grasping the battlefield dynamics for military applications, but is disturbed by low-resolution and noisyrepresentation. In this paper, a recognition method, involving a novel visual attention mechanismbased Gabor region proposal sub-network(Gabor RPN) and improved refinement generative adversarial sub-network(GAN), is proposed. Novel central-peripheral rivalry 3D color Gabor filters are proposed to simulate retinal structures and taken as feature extraction convolutional kernels in low-level layer to improve the recognition accuracy and framework training efficiency in Gabor RPN. Improved refinement GAN is used to solve the problem of blurry target classification, involving a generator to directly generate large high-resolution images from small blurry ones and a discriminator to distinguish not only real images vs. fake images but also the class of targets. A special recognition dataset for ground military target, named Ground Military Target Dataset(GMTD), is constructed. Experiments performed on the GMTD dataset effectively demonstrate that our method can achieve better energy-saving and recognition results when low-resolution and noisy-representation targets are involved, thus ensuring this algorithm a good engineering application prospect.展开更多
Owing to the wide range of applications in various fields,generative models have become increasingly popular.However,they do not handle spatio-temporal features well.Inspired by the recent advances in these models,thi...Owing to the wide range of applications in various fields,generative models have become increasingly popular.However,they do not handle spatio-temporal features well.Inspired by the recent advances in these models,this paper designs a distributed spatio-temporal generative adversarial network(STGAN-D)that,given some initial data and random noise,generates a consecutive sequence of spatio-temporal samples which have a logical relationship.This paper builds a spatio-temporal discriminator to distinguish whether the samples generated by the generator meet the requirements for time and space coherence,and builds a controller for distributed training of the network gradient updated to separate the model training and parameter updating,to improve the network training rate.The model is trained on the skeletal dataset and the traffic dataset.In contrast to traditional generative adversarial networks(GANs),the proposed STGAN-D can generate logically coherent samples with the corresponding spatial and temporal features while avoiding mode collapse.In addition,this paper shows that the proposed model can generate different styles of spatio-temporal samples given different random noise inputs,and the controller can improve the network training rate.This model will extend the potential range of applications of GANs to areas such as traffic information simulation and multiagent adversarial simulation.展开更多
心脏磁共振成像(cardiac magnetic resonance,CMR)过程中患者误动、异常幅度的呼吸运动、心律失常会造成CMR图像质量下降,为解决现有的CMR图像增强网络需要人为制作配对数据,且图像增强后部分组织纹理细节丢失的问题,提出了基于空频域...心脏磁共振成像(cardiac magnetic resonance,CMR)过程中患者误动、异常幅度的呼吸运动、心律失常会造成CMR图像质量下降,为解决现有的CMR图像增强网络需要人为制作配对数据,且图像增强后部分组织纹理细节丢失的问题,提出了基于空频域特征学习的循环一致性生成对抗网络(cycle-consistent generative adversavial network based on spatial-frequency domain feature learning,SFFL-CycleGAN).研究结果表明,该网络无须人为制作配对数据集,增强后的CMR图像组织纹理细节丰富,在结构相似度(structural similarity,SSIM)和峰值信噪比(peak signal to noise ratio,PSNR)等方面均优于现有的配对训练网络以及原始的CycleGAN网络,图像增强效果好,有效助力病情诊断.展开更多
基金This research was funded by Natural Science Foundation of Jiangsu Province,grant number BK20180579.
文摘The method of describing deformation camouflage spots based on feature space has some shortcomings,such as inaccurate description and difficult reproduction.Depending on the strong fitting ability of the generative adversarial network model,the distribution of deformation camouflage spot pattern can be directly fitted,thus simplifying the process of spot extraction and reproduction.The requirements of background spot extraction are analyzed theoretically.The calculation formula of limiting the range of image spot pixels is given and two kinds of spot data sets,forestland and snowfield,are established.Spot feature is decomposed into shape,size and color features,and a GAN(Generative Adversarial Network)framework is established.The effects of different loss functions on network training results are analyzed in the experiment.In the meantime,when the input dimension of generator network is 128,the balance between sample diversity and quality can be achieved.The effects of sample generation are investigated in two aspects.Subjectively,the probability of the generated spots being distinguished in the background is counted,and the results are all less than 20% and mostly close to zero.Objectively,the features of the spot shape are calculated and the independent sample T-test is applied to verify that the features are from the same distribution,and all the P-Values are much higher than 0.05.Both subjective and objective methods prove that the spots generated by this method are similar to the background spots.The proposed method can directly generate the desired camouflage pattern spots,which provides a new technical method for the deformation camouflage pattern design and camouflage effect evaluation.
基金supported by the Engineering and Physical Sciences Research Council [grant number: EP/N509644/1]。
文摘It is important to understand how ballistic materials respond to impact from projectiles such that informed decisions can be made in the design process of protective armour systems. Ballistic testing is a standards-based process where materials are tested to determine whether they meet protection, safety and performance criteria. For the V50ballistic test, projectiles are fired at different velocities to determine a key design parameter known as the ballistic limit velocity(BLV), the velocity above which projectiles perforate the target. These tests, however, are destructive by nature and as such there can be considerable associated costs, especially when studying complex armour materials and systems. This study proposes a unique solution to the problem using a recent class of machine learning system known as the Generative Adversarial Network(GAN). The GAN can be used to generate new ballistic samples as opposed to performing additional destructive experiments. A GAN network architecture is tested and trained on three different ballistic data sets, and their performance is compared. The trained networks were able to successfully produce ballistic curves with an overall RMSE of between 10 and 20 % and predicted the V50BLV in each case with an error of less than 5 %. The results demonstrate that it is possible to train generative networks on a limited number of ballistic samples and use the trained network to generate many new samples representative of the data that it was trained on. The paper spotlights the benefits that generative networks can bring to ballistic applications and provides an alternative to expensive testing during the early stages of the design process.
基金the National Key Research and Development Program of China(No.2016YFC0802904)National Natural Science Foundation of China(No.61671470)Natural Science Foundation of Jiangsu Province(BK20161470).
文摘Ground military target recognition plays a crucial role in unmanned equipment and grasping the battlefield dynamics for military applications, but is disturbed by low-resolution and noisyrepresentation. In this paper, a recognition method, involving a novel visual attention mechanismbased Gabor region proposal sub-network(Gabor RPN) and improved refinement generative adversarial sub-network(GAN), is proposed. Novel central-peripheral rivalry 3D color Gabor filters are proposed to simulate retinal structures and taken as feature extraction convolutional kernels in low-level layer to improve the recognition accuracy and framework training efficiency in Gabor RPN. Improved refinement GAN is used to solve the problem of blurry target classification, involving a generator to directly generate large high-resolution images from small blurry ones and a discriminator to distinguish not only real images vs. fake images but also the class of targets. A special recognition dataset for ground military target, named Ground Military Target Dataset(GMTD), is constructed. Experiments performed on the GMTD dataset effectively demonstrate that our method can achieve better energy-saving and recognition results when low-resolution and noisy-representation targets are involved, thus ensuring this algorithm a good engineering application prospect.
基金the National Natural Science Foundation of China(61573285).
文摘Owing to the wide range of applications in various fields,generative models have become increasingly popular.However,they do not handle spatio-temporal features well.Inspired by the recent advances in these models,this paper designs a distributed spatio-temporal generative adversarial network(STGAN-D)that,given some initial data and random noise,generates a consecutive sequence of spatio-temporal samples which have a logical relationship.This paper builds a spatio-temporal discriminator to distinguish whether the samples generated by the generator meet the requirements for time and space coherence,and builds a controller for distributed training of the network gradient updated to separate the model training and parameter updating,to improve the network training rate.The model is trained on the skeletal dataset and the traffic dataset.In contrast to traditional generative adversarial networks(GANs),the proposed STGAN-D can generate logically coherent samples with the corresponding spatial and temporal features while avoiding mode collapse.In addition,this paper shows that the proposed model can generate different styles of spatio-temporal samples given different random noise inputs,and the controller can improve the network training rate.This model will extend the potential range of applications of GANs to areas such as traffic information simulation and multiagent adversarial simulation.
文摘心脏磁共振成像(cardiac magnetic resonance,CMR)过程中患者误动、异常幅度的呼吸运动、心律失常会造成CMR图像质量下降,为解决现有的CMR图像增强网络需要人为制作配对数据,且图像增强后部分组织纹理细节丢失的问题,提出了基于空频域特征学习的循环一致性生成对抗网络(cycle-consistent generative adversavial network based on spatial-frequency domain feature learning,SFFL-CycleGAN).研究结果表明,该网络无须人为制作配对数据集,增强后的CMR图像组织纹理细节丰富,在结构相似度(structural similarity,SSIM)和峰值信噪比(peak signal to noise ratio,PSNR)等方面均优于现有的配对训练网络以及原始的CycleGAN网络,图像增强效果好,有效助力病情诊断.
