- 标题
- 摘要
- 关键词
- 实验方案
- 产品
-
Development of Limited-Angle Iterative Reconstruction Algorithms with Context Encoder-Based Sinogram Completion for Micro-CT Applications
摘要： Limited-angle iterative reconstruction (LAIR) reduces the radiation dose required for computed tomography (CT) imaging by decreasing the range of the projection angle. We developed an image-quality-based stopping-criteria method with a flexible and innovative instrument design that, when combined with LAIR, provides the image quality of a conventional CT system. This study describes the construction of different scan acquisition protocols for micro-CT system applications. Fully-sampled Feldkamp (FDK)-reconstructed images were used as references for comparison to assess the image quality produced by these tested protocols. The insufficient portions of a sinogram were inpainted by applying a context encoder (CE), a type of generative adversarial network, to the LAIR process. The context image was passed through an encoder to identify features that were connected to the decoder using a channel-wise fully-connected layer. Our results evidence the excellent performance of this novel approach. Even when we reduce the radiation dose by 1/4, the iterative-based LAIR improved the full-width half-maximum, contrast-to-noise and signal-to-noise ratios by 20% to 40% compared to a fully-sampled FDK-based reconstruction. Our data support that this CE-based sinogram completion method enhances the efficacy and efficiency of LAIR and that would allow feasibility of limited angle reconstruction.
关键词： generative adversarial network (GAN),context encoder (CE),limited-angle iterative reconstruction (LAIR)
更新于2025-09-19 17:15:36
-
Self-Mixing Interferometric Signal Enhancement Using Generative Adversarial Network for Laser Metric Sensing Applications
摘要： Measurement performance of self-mixing interferometric (SMI) laser sensor can be significantly affected due to the presence of noise. In this case, conventional signal enhancement techniques yield compromised performance due to several limitations which include processing signals in frequency domains only, relying mainly on first order statistics, loss of important information present in higher frequency band and handling limited number of noise types. To address these issues, we propose a solution based on using generative adversarial network, a popular deep learning scheme, to enhance SMI signal corrupted with different noise types. Thus, taking advantage of the deep networks that can learn arbitrary noise distribution from large example set, our proposed method trains the deep network model end-to-end, able to process raw waveforms directly, learn 51 different noise conditions including white noise and amplitude modulation noise for 1,140 different types of SMI waveforms made up of 285 different optical feedback coupling factor (C) values and 4 different line-width enhancement factor α values. The results show that the proposed method is able to significantly improve the SNR of noisy SM signals on average of 19.49, 16.29, 10.34 dB for weak-, moderate-, and strong-optical feedback regime signals, respectively. For amplitude modulated SMI signals, the proposed method has corrected the amplitude modulation with maximum error (using area-under-the-curve based quantitative analysis) of 0.73% for SMI signals belonging to all optical feedback regimes. Thus, our proposed method can effectively reduce the noise without distorting the original signal. We believe that such a unified and precise method leads to enhancement of performance of SMI laser sensors operating under real-world, noisy conditions.
关键词： signal noise removal,neural network for signal enhancement,Interferometry laser sensors,vibration measuring laser sensors,waveform enhancement,self-mixing signal enhancement,generative adversarial network (GAN)
更新于2025-09-16 10:30:52
-
Unsupervised Feature Extraction in Hyperspectral Images Based on Wasserstein Generative Adversarial Network
摘要： Feature extraction (FE) is a crucial research area in hyperspectral image (HSI) processing. Recently, due to the powerful ability of deep learning (DL) to extract spatial and spectral features, DL-based FE methods have shown great potentials for HSI processing. However, most of the DL-based FE methods are supervised, and the training of them suffers from the absence of labeled samples in HSIs severely. The training issue of supervised DL-based FE methods limits their application on HSI processing. To address this issue, in this paper, a novel modified generative adversarial network (GAN) is proposed to train a DL-based feature extractor without supervision. The designed GAN consists of two components, which are a generator and a discriminator. The generator can focus on the learning of real probability distributions of data sets and the discriminator can extract spatial–spectral features with superior invariance effectively. In order to learn upsampling and downsampling strategies adaptively during FE, the proposed generator and discriminator are designed based on a fully deconvolutional subnetwork and a fully convolutional subnetwork, respectively. Moreover, a novel min–max cost function is designed for training the proposed GAN in an end-to-end fashion without supervision, by utilizing the zero-sum game relationship between the generator and discriminator. Besides, the proposed modified GAN replaces the original Jensen–Shannon divergence with the Wasserstein distance, aiming to mitigate the unstability and difficulty of the training of GAN frameworks. Experimental results on three real data sets validate the effectiveness of the proposed method.
关键词： Convolutional neural network (CNN),hyperspectral images (HSIs),feature extraction (FE),generative adversarial network (GAN)
更新于2025-09-11 14:15:04
-
[IEEE 2019 International Russian Automation Conference - Sochi, Russia (2019.9.8-2019.9.14)] 2019 International Russian Automation Conference (RusAutoCon) - Laser Engraver Control System based on Reinforcement Adversarial Learning
摘要： In recent years, deep neural networks have demonstrated high efficiency in solving the problems associated with image processing and analysis. The latest studies in generative adversarial neural networks open up broad prospects for solving problems in image style transfer, cross-domain adaptation, and the generation of new data for a target probability distribution. The ability of generative networks to implement complex and weakly formalized transformations allows the information systems based on generative adversarial technologies to solve problems in machine learning in a manner similar to human creative thinking. This paper presents the results of an experimental study on the design, training and deployment of an intelligent control system for a CNC laser engraver based on the generative adversarial network. In order to train deep models, large amounts of data are required. Therefore, one of the main challenges in this experimental study is a lack of labelled training dataset. The proposed control system for the laser engraver based on the deep generative model allows decreasing the time of the manufacturing process.
关键词： generative adversarial network,intelligent control system,deep learning,reinforcement learning
更新于2025-09-11 14:15:04
-
[IEEE IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium - Valencia, Spain (2018.7.22-2018.7.27)] IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium - Multi-Discriminator Generative Adversarial Network for High Resolution Gray-Scale Satellite Image Colorization
摘要： Automatic colorization for gray-scale satellite images can help with eliminating lighting differences between multi-spectral captures, and provides strong prior information for ground type classification and object detection. In this paper, we introduced a novel generative adversarial network with multiple discriminators for colorizing gray-scale satellite images with pseudo-natural appearances. Although being powerful, deep generative model in its common form with a single discriminator could be unstable for achieving spatial consistency on local textured regions, especially highly textured ones. To address this issue, the generator in our proposed structure produces a group of colored outputs from feature maps at different scale levels of the network, each being supervised by an independent discriminator to fit the original colored training input in discrete Lab color space. The final colored output is a cascaded ensemble of these preceding by-products via summation, thus the fitting errors are reduced by a geometric series form. Quantitative and qualitative comparisons with the sole-discriminator version have been performed on high-resolution satellite images in experiments, where significant reductions in prediction errors have been observed.
关键词： gray-scale satellite images,generative adversarial network,Pseudo-natural colorization,multiple discriminators
更新于2025-09-10 09:29:36
-
[IEEE IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium - Valencia (2018.7.22-2018.7.27)] IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium - Deep Generative Matching Network for Optical and SAR Image Registration
摘要： Multimodal remote sensing images contain complementary information, thus, could potentially benefit many remote sensing applications. To this end, the image registration is a common requirement for utilizing the multimodal images. However, due to the rather different imaging mechanisms, multimodal image registration becomes much more challenging than ordinary registration, particular for optical and synthetic aperture radar (SAR) images. In this work, we design a deep matching network to exploit the latent and coherent features between multimodal patch pairs for inferring their matching labels. But, the network requires immense data for training, which is not usually met. To address this issue, we propose a generative matching network (GMN) to generate the coupled optical and SAR images, hence, improve the quantity and diversity of the training data. The experimental results show that our proposal significantly improves the registration performance of optical and SAR image registration, and achieves subpixel or close to subpixel error.
关键词： multimodal images,generative adversarial network,optical and SAR,deep matching network,image registration
更新于2025-09-09 09:28:46
-
[IEEE 2018 25th IEEE International Conference on Image Processing (ICIP) - Athens, Greece (2018.10.7-2018.10.10)] 2018 25th IEEE International Conference on Image Processing (ICIP) - Near InfraRed Imagery Colorization
摘要： This paper proposes a stacked conditional Generative Adversarial Network-based method for Near InfraRed (NIR) imagery colorization. We propose a variant architecture of Generative Adversarial Network (GAN) that uses multiple loss functions over a conditional probabilistic generative model. We show that this new architecture/loss-function yields better generalization and representation of the generated colored IR images. The proposed approach is evaluated on a large test dataset and compared to recent state of art methods using standard metrics.
关键词： Convolutional Neural Networks (CNN),Infrared Imagery colorization,Generative Adversarial Network (GAN)
更新于2025-09-04 15:30:14
-
[IEEE 2018 26th International Conference on Geoinformatics - Kunming, China (2018.6.28-2018.6.30)] 2018 26th International Conference on Geoinformatics - Generative Adversarial Network for Deblurring of Remote Sensing Image
摘要： Deblurring is a classical problem for remote sensing images, which is known to be difficult as an ill-posed problem. A feasible solution for the problem is incorporating various priors into restoration procedure as constrained conditions. However, the learning of priors usually assumes that the blurs in an image are produced by fixed types of reasons, and thus a possible decrease in model’s description ability. In this paper, an end-to-end learned method based on generative adversarial networks (GANs) is proposed to tackle the deblurring problem for remote sensing images. The proposed deblurring model does not need any prior assumptions for the blurs. The proposed method was evaluated on a satellite map image data set and state-of-the-art performance was obtained.
关键词： image deblurring,remote sensing image,loss function,Generative Adversarial Network (GAN)
更新于2025-09-04 15:30:14
-
[IEEE IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium - Valencia (2018.7.22-2018.7.27)] IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium - Can We Generate Good Samples for Hyperspectral Classification? — A Generative Adversarial Network Based Method
摘要： The insufficiency of training samples is really a great challenge for hyperspectral image (HSI) classification. Samples generation is a commonly used technique in deep learning based remote sensing field which can extend the training set. However, previous methods ignore the real distribution of the training samples in the feature space and thus can hardly ensure that the generated samples possess the same patterns with the real ones. In this paper, we propose a generative adversarial network based method (SpecGAN) to handle this problem. Different from traditional GAN framework where the generated samples have no categories, for the first time we take the label information into consideration for hyperspectral images. Feeding a random noise z and a class label vector y into the generator, we can get a spectral sample of the corresponding category. The experiments on the Pavia University data set demonstrate the potential of the proposed SpecGAN in spectral samples generation.
关键词： hyperspectral image classification,generative adversarial network,Sample generation,deep learning
更新于2025-09-04 15:30:14