Filters








1,250 Hits in 5.0 sec

Cross-Domain Medical Image Translation by Shared Latent Gaussian Mixture Model [article]

Yingying Zhu, Youbao Tang, Yuxing Tang, Daniel C. Elton, Sungwon Lee, Perry J. Pickhardt, Ronald M. Summers
2020 arXiv   pre-print
In order to preserve fine structures during medical image translation, we propose a patch-based model using shared latent variables from a Gaussian mixture model.  ...  Many existing cross-domain image-to-image translation models have been shown to improve cross-domain segmentation of large organs.  ...  Acknowledgments This research was supported in part by the Intramural Research Program of the National Institutes of Health Clinical Center. We thank NVIDIA for GPU card donations.  ... 
arXiv:2007.07230v1 fatcat:3iw3cr6rnjhqvf2mjj5ecezjpm

Deep Generative Adversarial Networks for Image-to-Image Translation: A Review

Aziz Alotaibi
2020 Symmetry  
Image-to-image translation with generative adversarial networks (GANs) has been intensively studied and applied to various tasks, such as multimodal image-to-image translation, super-resolution translation  ...  Many image processing, computer graphics, and computer vision problems can be treated as image-to-image translation tasks.  ...  The attribute latent space was modeled by a Gaussian mixture model (GMM); thus, the model was named GMM-UNIT. In this model, each Gaussian component in a mixture was associated with a domain.  ... 
doi:10.3390/sym12101705 fatcat:rqlwjjhrvbc6fhc4mxjjvkwk6i

True-false lumen segmentation of aortic dissection using multi-scale wavelet analysis and generative-discriminative model matching

Noah Lee, Huseyin Tek, Andrew F. Laine, Maryellen L. Giger, Nico Karssemeijer
2008 Medical Imaging 2008: Computer-Aided Diagnosis  
Computer aided diagnosis in the medical image domain requires sophisticated probabilistic models to formulate quantitative behavior in image space.  ...  In our approach, we perform probabilistic model matching of both modeling domains to fuse into the prediction step appearance and structural information of the object of interest while exploiting the strength  ...  ACKNOWLEDGEMENTS This work was supported by the Imaging and Visualization Department at Siemens Corporate Research.  ... 
doi:10.1117/12.770610 dblp:conf/micad/LeeTL08 fatcat:klpyzlpkgjehlbopupeh7bcgg4

Variational Resampling Based Assessment of Deep Neural Networks under Distribution Shift [article]

Xudong Sun, Alexej Gossmann, Yu Wang, Bernd Bischl
2019 arXiv   pre-print
We use Auto Encoding Variational Bayes to find a latent representation of the data, on which a Variational Gaussian Mixture Model is applied to deliberately create distribution shift by dividing the dataset  ...  We compare several popular Convolutional Neural Network (CNN) architectures and Bayesian CNN models for image classification on the Fashion-MNIST dataset, to assess their robustness and generalization  ...  VARITIONAL GAUSSIAN MIXTURE MODELS ON THE DATA LATENT SPACE TABLE IV PAIRWISE IV WASSERSTEIN DISTANCE ACROSS 5 CLUSTERS CREATED BY RANDOM SPLITTING OF THE DATA  ... 
arXiv:1906.02972v6 fatcat:w4n4mb2zubatjocl6lkq5simeu

Image-to-Image Translation: Methods and Applications [article]

Yingxue Pang, Jianxin Lin, Tao Qin, Zhibo Chen
2021 arXiv   pre-print
Image-to-image translation (I2I) aims to transfer images from a source domain to a target domain while preserving the content representations.  ...  I2I has drawn increasing attention and made tremendous progress in recent years because of its wide range of applications in many computer vision and image processing problems, such as image synthesis,  ...  [74] propose UNIT to make a shared latent space assumption that a pair of corresponding images in different domains can be mapped to the same latent code in a shared latent space.  ... 
arXiv:2101.08629v2 fatcat:i6pywjwnvnhp3i7cmgza2slnle

Learning Disentangled Representations in the Imaging Domain [article]

Xiao Liu, Pedro Sanchez, Spyridon Thermos, Alison Q. O'Neil, Sotirios A. Tsaftaris
2021 arXiv   pre-print
We discuss applications in medical imaging and computer vision emphasising choices made in exemplar key works. We conclude by presenting remaining challenges and opportunities.  ...  A good general representation can be fine-tuned for new target tasks using modest amounts of data, or used directly in unseen domains achieving remarkable performance in the corresponding task.  ...  Acknowledgments This work was supported by the Royal Academy of Engineering and Canon Medical Research Europe, and partially supported by the Alan Turing Institute under the EPSRC grant EP/N510129/1.  ... 
arXiv:2108.12043v4 fatcat:rckcevnvw5apdnslkr76a65ztu

Unsupervised Domain Adaptation with Variational Approximation for Cardiac Segmentation [article]

Fuping Wu, Xiahai Zhuang
2021 arXiv   pre-print
Unsupervised domain adaptation is useful in medical image segmentation.  ...  Particularly, when ground truths of the target images are not available, domain adaptation can train a target-specific model by utilizing the existing labeled images from other modalities.  ...  For segmentation, the VAE incorporates a prediction model, which takes only latent features as input, and thus can be shared by both domains.  ... 
arXiv:2106.08752v1 fatcat:yplkd6yvtnd7vhy6luozum3ctm

A Probabilistic U-Net for Segmentation of Ambiguous Images [article]

Simon A. A. Kohl, Bernardino Romera-Paredes, Clemens Meyer, Jeffrey De Fauw, Joseph R. Ledsam, Klaus H. Maier-Hein, S. M. Ali Eslami, Danilo Jimenez Rezende, Olaf Ronneberger
2019 arXiv   pre-print
To this end we propose a generative segmentation model based on a combination of a U-Net with a conditional variational autoencoder that is capable of efficiently producing an unlimited number of plausible  ...  These models could have a high impact in real-world applications, such as being used as clinical decision-making algorithms accounting for multiple plausible semantic segmentation hypotheses to provide  ...  There however inevitably are transitions between those latent space regions that will translate to mixtures of the grader modes that cross over.  ... 
arXiv:1806.05034v4 fatcat:xnnzsbwvm5actekgwrlskpnld4

How Generative Adversarial Networks and Their Variants Work

Yongjun Hong, Uiwon Hwang, Jaeyoon Yoo, Sungroh Yoon
2019 ACM Computing Surveys  
This powerful property leads GAN to be applied to various applications such as image synthesis, image attribute editing, image translation, domain adaptation and other academic fields.  ...  Specifically, they do not rely on any assumptions about the distribution and can generate real-like samples from latent space in a simple manner.  ...  Image Image translation Image translation involves translating images in one domain X to images in another domain Y .  ... 
doi:10.1145/3301282 fatcat:z2xe6jdh5nd2dmovkes3rav3ke

Log-Likelihood Ratio Minimizing Flows: Towards Robust and Quantifiable Neural Distribution Alignment [article]

Ben Usman, Avneesh Sud, Nick Dufour, Kate Saenko
2020 arXiv   pre-print
Distribution alignment has many applications in deep learning, including domain adaptation and unsupervised image-to-image translation.  ...  We experimentally verify that minimizing the resulting objective results in domain alignment that preserves the local structure of input domains.  ...  Unfortunately, in many important problem domains, such as medical imaging, labeling costs and high variability of target environments, such as differences in image capturing medical equipment, prohibit  ... 
arXiv:2003.12170v2 fatcat:fmjiiub7gndctg4fqnbjxiko2i

Conditional Generation of Medical Images via Disentangled Adversarial Inference [article]

Mohammad Havaei, Ximeng Mao, Yiping Wang, Qicheng Lao
2021 arXiv   pre-print
We show that in general, two latent variable models achieve better performance and give more control over the generated image.  ...  First, we minimize the shared information between content and style by introducing a novel application of the gradient reverse layer (GRL); second, we introduce a self-supervised regularization method  ...  The cycle-consistency component is used for indomain reconstruction and the adversarial component is used for cross-domain translation.  ... 
arXiv:2012.04764v2 fatcat:kikndmwg5ng2hjxyfumuouih6a

Disentanglement by Nonlinear ICA with General Incompressible-flow Networks (GIN) [article]

Peter Sorrenson, Carsten Rother, Ullrich Köthe
2020 arXiv   pre-print
from noise by an estimating model.  ...  Recent breakthrough work by Khemakhem et al. (2019) on nonlinear ICA has answered this question for a broad class of conditional generative processes.  ...  Variational autoencoder with truncated mixture of gaussians for functional connectivity analysis. In International Conference on Information Processing in Medical Imaging, pp. 867- 879.  ... 
arXiv:2001.04872v1 fatcat:epijjewfgfeslbf66hlwuwyeny

Integration of Unpaired Single-cell Chromatin Accessibility and Gene Expression Data via Adversarial Learning [article]

Yang Xu, Andrew Jeremiah Strick
2021 arXiv   pre-print
We demonstrate that our method substantially improves data integration from a simple adversarial domain adaption approach, and it also outperforms two state-of-the-art (SOTA) methods.  ...  Starting from the original image, a generator network translates the image to the other domain. Then, a second generator network translates the image back to its original domain.  ...  Generative models with adversarial domain adaption were successfully shown to transfer targets to source style and has its application in image translation [22] .  ... 
arXiv:2104.12320v1 fatcat:sb6cokdqpbhm5cyu3nxcgo5f6a

Disentangle, align and fuse for multimodal and semi-supervised image segmentation [article]

Agisilaos Chartsias, Giorgos Papanastasiou, Chengjia Wang, Scott Semple, David E. Newby, Rohan Dharmakumar, Sotirios A. Tsaftaris
2020 arXiv   pre-print
We present a method that offers improved segmentation accuracy of the modality of interest (over a single input model), by learning to leverage information present in other modalities, even if few (semi-supervised  ...  Shared anatomical factors from the different inputs are jointly processed and fused to extract more accurate segmentation masks.  ...  The first uses MUNIT to translate images from source to target modality [9] , and the second translates multimodal images to a domain invariant space [51] .  ... 
arXiv:1911.04417v4 fatcat:qxlay6fzz5fdlcpta2epygydf4

Generative Adversarial Network in Medical Imaging: A Review [article]

Xin Yi, Ekta Walia, Paul Babyn
2019 arXiv   pre-print
This has proven to be useful in many cases, such as domain adaptation, data augmentation, and image-to-image translation.  ...  , classification, and cross-modality synthesis.  ...  Another model named UNIT (Figure 4 c) can also perform unpaired image-to-image transform by combining two VAEGANs together with each one responsible for one modality but sharing the same latent space  ... 
arXiv:1809.07294v3 fatcat:5j5i6shlcvbbjm74ceidzg6rc4
« Previous Showing results 1 — 15 out of 1,250 results