Filters








21,070 Hits in 6.1 sec

Removal of Batch Effects using Generative Adversarial Networks [article]

Uddeshya Upadhyay, Arjun Jain
2019 arXiv   pre-print
This can be solved using a novel Generative Adversarial Networks (GANs) based framework that is proposed here, advantage of using this framework over other prior approaches is that here it is not required  ...  Many biological data analysis processes like Cytometry or Next Generation Sequencing (NGS) produce massive amounts of data which needs to be processed in batches for down-stream analysis.  ...  Results Conclusions We present a novel solution to correct for batch effects which are very frequent in biological data analysis using Generative Adversarial Networks.  ... 
arXiv:1901.06654v3 fatcat:3fzg6a2i3jdyrnbzzt2zgwvtay

Intriguing properties of adversarial training at scale [article]

Cihang Xie, Alan Yuille
2019 arXiv   pre-print
Batch normalization (BN) is a crucial element for achieving state-of-the-art performance on many vision tasks, but we show it may prevent networks from obtaining strong robustness in adversarial training  ...  Second, we study the role of network capacity. We find our so-called "deep" networks are still shallow for the task of adversarial learning.  ...  As a side note, this MBN structure is also used as a practical trick for training better generative adversarial networks (GAN) .  ... 
arXiv:1906.03787v2 fatcat:3uu6sbnvljbhnlm6xsy4fy3uvi

Entropy Guided Adversarial Model for Weakly Supervised Object Localization [article]

Sabrina Narimene Benassou, Wuzhen Shi, Feng Jiang
2020 arXiv   pre-print
In this present article, we propose to take advantage of the generalization ability of the network and train the model using clean examples and adversarial examples to localize the whole object.  ...  Some methods tend to remove some parts of the object to force the CNN to detect other features, whereas, others change the network structure to generate multiple CAMs from different levels of the model  ...  Specifically, for each clean mini batch x clean , we generate its corresponding adversarial mini-batch x adv using the auxiliary batch norm layer.  ... 
arXiv:2008.01786v1 fatcat:4gcz5xer7zbrlah24xeg5jau6u

Image Rain Removal Using Conditional Generative Networks Incorporating

Fangyan Zhang, Xinzheng Xu, Peng Wang
2022 Journal of Computer and Communications  
We exploit the powerful generative power of a modified generative adversarial network (CGAN) by enforcing an additional condition that makes the derained image indistinguishable from its corresponding  ...  Most noise reduction methods more or less remove texture details in rain-free areas, resulting in an over-smoothing effect in the restored background.  ...  But it is not very friendly to the details of rain removal. In this paper, we study the effectiveness of conditional generative adversarial network (GAN) in solving this problem.  ... 
doi:10.4236/jcc.2022.102006 fatcat:z3k3gttlorcvheniatgomazigm

Adversarial Deconfounding Autoencoder for Learning Robust Gene Expression Embeddings [article]

Ayse B. Dincer, Joseph D. Janizek, Su-In Lee
2020 bioRxiv   pre-print
., batch effects) and uninteresting biological variables (e.g., age) in addition to the true signals of interest.  ...  Motivation: Increasing number of gene expression profiles has enabled the use of complex models, such as deep unsupervised neural networks, to extract a latent space from these profiles.  ...  We also thankfully acknowledge all members of the AIMS lab for their helpful comments and useful discussions.  ... 
doi:10.1101/2020.04.28.065052 fatcat:yojdc3lp6jcf3bguv7g2zjeg2u

A New Image Denoising Method with Gan Models

Xueji Huang
2020 Journal of Electronic Research and Application  
There are residual noise features in the obtained denoising features, which are removed by subsequent feature filtering of the network structure, and finally a denoised image is generated by fusing the  ...  This paper proposes a q-GAN, which uses multi-scale in generating networks. The convolution kernel extracts image features and transforms the denoising problem into the feature domain.  ...  The structure of discriminant network and the advantages of adversarial training are explained. The loss function of the generative adversarial network is given.  ... 
doi:10.26689/jera.v4i2.1159 fatcat:eqqwxss6mbhwtlcxukttc44sxq

Towards an Adversarially Robust Normalization Approach [article]

Muhammad Awais, Fahad Shamshad, Sung-Ho Bae
2020 arXiv   pre-print
Batch Normalization (BatchNorm) is effective for improving the performance and accelerating the training of deep neural networks.  ...  However, without estimated batch statistics, we can not use BatchNorm in the practice if large batches of input are not available.  ...  Table shows effect of using population vs batch statis- tics on adversarial accuracy for several models on CIFAR10 and Imagenet.  ... 
arXiv:2006.11007v1 fatcat:c4qim5ph2jfhjcnhjic4okiooa

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

Yang Xu, Andrew Jeremiah Strick
2021 arXiv   pre-print
Deep learning has empowered analysis for single-cell sequencing data in many ways and has generated deep understanding about a range of complex cellular systems.  ...  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.  ...  Most batch effects are not biologically relevant, and single-cell database confounded by batch effects are also not applicable for general use.  ... 
arXiv:2104.12320v1 fatcat:sb6cokdqpbhm5cyu3nxcgo5f6a

Deep feature extraction of single-cell transcriptomes by generative adversarial network [article]

Mojtaba Bahrami, Malosree Maitra, Corina Nagy, Gustavo Turecki, Hamid Rabiee, Yue Li
2020 bioRxiv   pre-print
Our main contribution is to introduce an adversarial network to predict batch effects using the embeddings from the variational autoencoder network, which does not only need to maximize the Negative Binomial  ...  However, batch effects such as laboratory conditions and individual-variability hinder their usage in cross-condition design. We present single-cell Generative Adversarial Network (scGAN).  ...  Conclusion In summary, we demonstrate the utility of attenuating batch effects via an adversarial network while learning the low-dimensional single-cell embedding from the high dimensional scRNAseq profiles  ... 
doi:10.1101/2020.04.29.066464 fatcat:ebnojkznl5b7bccfq6un7vcb3m

Fast AdvProp [article]

Jieru Mei, Yucheng Han, Yutong Bai, Yixiao Zhang, Yingwei Li, Xianhang Li, Alan Yuille, Cihang Xie
2022 arXiv   pre-print
Adversarial Propagation (AdvProp) is an effective way to improve recognition models, leveraging adversarial examples.  ...  their adversarial counterparts are used for training (i.e., 2× data).  ...  Adversarial training (Szegedy et al., 2014; Goodfellow et al., 2015) , which trains networks with adversarial examples that are generated on the fly, is one of the most effective ways for defending against  ... 
arXiv:2204.09838v1 fatcat:at2qu5zyjvhvdm6ucanjos2a64

Generator From Edges: Reconstruction of Facial Images [article]

Nao Takano, Gita Alaghband
2020 arXiv   pre-print
First, we explored the effects of the adversarial loss often used in SISR. In particular, we uncovered that it is not an essential component to form a perceptual loss.  ...  Eliminating adversarial loss will lead to a more effective architecture from the perspective of hardware resource.  ...  • Experiment 1 -Effect of Adversarial Loss Using CelebA dataset, we measured the effectiveness of adversarial loss for both SISR and the synthesis from canny images.  ... 
arXiv:2002.06682v3 fatcat:lair7ooz6bhyfddtckqsmx4h4i

Cross Modality Microscopy Segmentation via Adversarial Adaptation [chapter]

Yue Guo, Qian Wang, Oleh Krupa, Jason Stein, Guorong Wu, Kira Bradford, Ashok Krishnamurthy
2019 Lecture Notes in Computer Science  
However, the training of deep learning networks requires a massive amount of manually-labeled training data, which is a very time-consuming operation.  ...  In this paper, we demonstrate an adversarial adaptation method to transfer deep network knowledge for microscopy segmentation from one imaging modality (e.g., confocal) to a new imaging modality (e.g.,  ...  [17] simplified CoGAN by removing the generative model of CoGAN and designed a discriminative model.  ... 
doi:10.1007/978-3-030-17935-9_42 pmid:32154516 pmcid:PMC7062366 fatcat:bmdankxwvjeyhgl2oxjdebhrty

Defense Against Adversarial Examples Using Quality Recovery for Image Classification

Motohiro TAKAGI, Masafumi HAGIWARA
2020 Journal of Japan Society for Fuzzy Theory and Intelligent Informatics  
To remove adversarial perturbation, the proposed network is trained using various types of distorted images considering the proposed PRloss metric.  ...  The proposed method, called the denoising-based perturbation removal network (DPRNet), aims to eliminate perturbations generated by an adversarial attack for image classification tasks.  ...  The accuracy of the distillation network was 0.7496 for the adversarial images generated by the C&W attack.  ... 
doi:10.3156/jsoft.32.4_811 fatcat:rktffmkdxje2bonveix34bfwdu

Advanced Single Image Resolution Upsurging Using A Generative Adversarial Network

Md. Moshiur Rahman, Samrat Kumar Dey, Kabid Hassan Shibly
2020 Signal & Image Processing An International Journal  
In this paper, we have proposed a technique of generating higher resolution images form lower resolution using Residual in Residual Dense Block network architecture with a deep network.  ...  In recent times, various research works are performed to generate a higher resolution of an image from its lower resolution.  ...  is depicted Here we are using GAN or Generative Adversarial Network.  ... 
doi:10.5121/sipij.2020.11105 fatcat:cn2cifwd5bcsfezjeopmet47ru

Adjusting for Confounding in Unsupervised Latent Representations of Images [article]

Craig A. Glastonbury, Michael Ferlaino, Christoffer Nellåker, Cecilia M. Lindgren
2018 arXiv   pre-print
Examples of such phenomena include variable lighting across microscopy image captures, stain intensity variation in histological slides, and batch effects for high throughput drug screening assays.  ...  In this paper, we present a strategy based on adversarial training, capable of learning unsupervised representations invariant to confounders.  ...  Acknowledgements We gratefully acknowledge the support of NVIDIA Corporation with the donation of 2 Titan Xp GPUs used for this research.  ... 
arXiv:1811.06498v2 fatcat:d7kewupucjaevjhd5pxqjfqb4e
« Previous Showing results 1 — 15 out of 21,070 results