基于生成对抗网络的红外图像数据增强
Infrared image data augmentation based on generative adversarial network
查看参考文献26篇
陈佛计
1,2,3,4
*
朱枫
1,2,3,4
吴清潇
1,2,3,4
郝颖明
1,2,3,4
王恩德
1,2,3,4
文摘
|
深度学习在视觉任务中的良好表现很大程度上依赖于海量的数据和计算力的提升,但是在很多实际项目中通常难以提供足够的数据来完成任务。针对某些情况下红外图像少且难以获得的问题,提出一种基于彩色图像生成红外图像的方法来获取更多的红外图像数据。首先,用现有的彩色图像和红外图像数据构建成对的数据集;然后,基于卷积神经网络、转置卷积神经网络构建生成对抗网络(GAN)模型的生成器和鉴别器;接着,基于成对的数据集来训练GAN模型,直到生成器和鉴别器之间达到纳什平衡状态;最后,用训练好的生成器将彩色图像从彩色域变换到红外域。基于定量评估标准对实验结果进行了评估,结果表明,所提方法可以生成高质量的红外图像,并且相较于在损失函数中不加正则化项,在损失函数中加入L1和L2正则化约束后,该方法的FID分数值平均分别降低了23.95和20.89。作为一种无监督的数据增强方法,该方法也可以被应用于其他缺少数据的目标识别、目标检测、数据不平衡等视觉任务中。 |
其他语种文摘
|
The great performance of deep learning in many visual tasks largely depends on the big data volume and the improvement of computing power. But in many practical projects,it is usually difficult to provide enough data to complete the task. Concerning the problem that the number of infrared images is small and the infrared images are hard to collect,a method to generate infrared images based on color images was proposed to obtain more infrared image data. Firstly,the existing color image and infrared image data were employed to construct the paired datasets. Secondly,the generator and the discriminator of Generative Adversarial Network(GAN)model were formed based on the convolutional neural network and the transposed convolutional neural network. Thirdly,the GAN model was trained based on the paired datasets until the Nash equilibrium between the generator and the discriminator was reached. Finally,the trained generator was used to transform the color image from the color field to the infrared field. The experimental results were evaluated based on quantitative evaluation metrics. The evaluation results show that the proposed method can generate high-quality infrared images. In addition,after the L1 or L2 regularization constraint was added to the loss function,the FID(Frechet Inception Distance)score was respectively reduced by 23.95,20.89 on average compared to the FID score of loss function not adding the constraint. As an unsupervised data augmentation method,the method can also be applied to many other visual tasks that lack train data,such as target recognition,target detection and data imbalance. |
来源
|
计算机应用
,2020,40(7):2084-2088 【扩展库】
|
DOI
|
10.11772/j.issn.1001-9081.2019122253
|
关键词
|
红外图像生成
;
生成对抗网络
;
图像转换
;
数据增强
;
生成图像质量评估
|
地址
|
1.
中国科学院沈阳自动化研究所, 沈阳, 110016
2.
中国科学院机器人与智能制造创新研究院, 沈阳, 110016
3.
中国科学院大学, 北京, 100049
4.
中国科学院光电信息处理重点实验室, 中国科学院光电信息处理重点实验室, 沈阳, 110016
|
语种
|
中文 |
文献类型
|
研究性论文 |
ISSN
|
1001-9081 |
学科
|
自动化技术、计算机技术 |
基金
|
国家自然科学基金资助项目
|
文献收藏号
|
CSCD:6766332
|
参考文献 共
26
共2页
|
1.
Dayan P. Helmholtz machines and wake-sleep learning.
Handbook of Brain Theory and Neural Network,2000:522-525
|
被引
1
次
|
|
|
|
2.
Hinton G E. Deep belief networks.
Scholarpedia,2009,4(5):5947
|
被引
34
次
|
|
|
|
3.
Kingma D P.
Auto-encoding variational Bayes,2019
|
被引
12
次
|
|
|
|
4.
Salakhutdinov R. Restricted Boltzmann machines for collaborative filtering.
Proceedings of the 24th International Conference on Machine Learning,2007:791-798
|
被引
48
次
|
|
|
|
5.
Salakhutdinov R. Deep Boltzmann machines.
Proceedings of the 2009 Artificial Intelligence and Statistics,2009:448-455
|
被引
1
次
|
|
|
|
6.
Van Den Oord A.
Pixel recurrent neural networks,2019
|
被引
2
次
|
|
|
|
7.
Goodfellow I. Generative adversarial nets.
Proceedings of the 27th International Conference on Neural Information Processing Systems,2014:2672-2680
|
被引
92
次
|
|
|
|
8.
林懿伦. 人工智能研究的新前线:生成式对抗网络.
自动化学报,2018,44(5):775-792
|
被引
49
次
|
|
|
|
9.
曹仰杰. 生成式对抗网络及其计算机视觉应用研究综述.
中国图象图形学报,2018,23(10):1433-1449
|
被引
23
次
|
|
|
|
10.
陈文兵. 基于条件生成式对抗网络的数据增强方法.
计算机应用,2018,38(11):3305-3311
|
被引
20
次
|
|
|
|
11.
Chen X. InfoGAN:interpretable representation learning by information maximizing generative adversarial nets.
Proceedings of the 30th International Conference on Neural Information Processing Systems,2016:2172-2180
|
被引
4
次
|
|
|
|
12.
Zhang H.
Self-attention generative adversarial networks,2019
|
被引
8
次
|
|
|
|
13.
Isola P. Image-to-image translation with conditional adversarial networks.
Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition,2017:5967-5976
|
被引
31
次
|
|
|
|
14.
Zhu J Y. Unpaired image-to-image translation using cycle-consistent adversarial networks.
Proceedings of the 2017 IEEE International Conference on Computer Vision,2017:2242-2251
|
被引
27
次
|
|
|
|
15.
Odena A. Conditional image synthesis with auxiliary classifier GANs.
Proceedings of the 34th International Conference on Machine Learning,2017:2642-2651
|
被引
27
次
|
|
|
|
16.
Choi Y. StarGAN:unified generative adversarial networks for multi-domain image-to-image translation.
Proceedings of the 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition,2018:8789-8797
|
被引
6
次
|
|
|
|
17.
许洪. 基于可见光光谱图像的红外多光谱图像仿真生成.
红外与激光工程,2009,38(2):200-204
|
被引
8
次
|
|
|
|
18.
陈珊. 基于可见光图像的红外场景仿真.
红外与激光工程,2009,38(1):23-26,30
|
被引
6
次
|
|
|
|
19.
Arjovsky M.
Wasserstein GAN,2019
|
被引
6
次
|
|
|
|
20.
Brock A.
Large scale GAN training for high fidelity natural image synthesis,2019
|
被引
12
次
|
|
|
|
|