22,222 Hits in 4.8 sec

A Bayesian Data Augmentation Approach for Learning Deep Models [article]

Toan Tran, Trung Pham, Gustavo Carneiro, Lyle Palmer, Ian Reid
2017 arXiv   pre-print
Data augmentation is an essential part of the training process applied to deep learning models.  ...  The motivation is that a robust training process for deep learning models depends on large annotated datasets, which are expensive to be acquired, stored and processed.  ...  Figure 1: An overview of our Bayesian data augmentation algorithm for learning deep models.  ... 
arXiv:1710.10564v1 fatcat:hvxo6htigvbovo26srofdnxwbe

Bayesian Generative Active Deep Learning [article]

Toan Tran, Thanh-Toan Do, Ian Reid, Gustavo Carneiro
2019 arXiv   pre-print
In this paper, we propose a Bayesian generative active deep learning approach that combines active learning with data augmentation -- we provide theoretical and empirical evidence (MNIST, CIFAR-{10,100  ...  Deep learning models have demonstrated outstanding performance in several problems, but their training process tends to require immense amounts of computational and human resources for training and labeling  ...  In this work we proposed a Bayesian generative active deep learning approach that consistently shows to be more effective than data augmentation and active learning in several classification problems.  ... 
arXiv:1904.11643v1 fatcat:xaw37dqjfzey3gtjgqge2e3vse

Sheep identification using a hybrid deep learning and Bayesian optimization approach

Aya Salama, Aboul Ellah Hassanien, Aly Fahmy.
2019 IEEE Access  
Our approach outperforms previous approaches for sheep identification. INDEX TERMS Bayesian optimization, convolutional neural network, deep learning.  ...  Thus, the data augmentation methodologies such as rotation, reflection, scaling, blurring, and brightness modification were applied; 1000 images of each sheep were obtained for training and validation.  ...  AUGMENTATION Deep learning approaches require a sufficient number of training images to boost their performance [20] .  ... 
doi:10.1109/access.2019.2902724 fatcat:5e5sj3vutfc7daoomlzoadfxti

BO-ALLCNN: Bayesian-Based Optimized CNN for Acute Lymphoblastic Leukemia Detection in Microscopic Blood Smear Images

Ghada Atteia, Amel A. Alhussan, Nagwan Abdel Samee
2022 Sensors  
The findings of this study reveal the superiority of the proposed Bayesian-optimized CNN over other optimized deep learning ALL classification models.  ...  Deep learning-based human-centric biomedical diagnosis has recently emerged as a powerful tool for assisting physicians in making medical decisions.  ...  Acknowledgments: We acknowledge Princess Nourah bint Abdulrahman University Researchers Supporting Project number (PNURSP2022R308), Princess Nourah bint Abdulrahman University, Riyadh, Saudi Arabia for  ... 
doi:10.3390/s22155520 pmid:35898023 pmcid:PMC9329984 fatcat:zml2hf73ofebjcxgklfgqnxaru

A voting-based ensemble deep learning method focusing on image augmentation and preprocessing variations for tuberculosis detection

Erdal Tasci, Caner Uluturk, Aybars Ugur
2021 Neural computing & applications (Print)  
image transformations for data augmentation types.  ...  We apply the Bayesian optimization-based weighted voting and the average of probabilities as a combination rule in soft voting methods on two TB CXR image datasets to get better results in various numbers  ...  Conclusion This study proposes a voting-based ensemble deep learning approach using diverse preprocessing/data augmentation variations on TB detection image datasets.  ... 
doi:10.1007/s00521-021-06177-2 pmid:34121816 pmcid:PMC8182991 fatcat:gilfochxj5glpkankw7xmwonwi

A Survey of Uncertainty in Deep Neural Networks [article]

Jakob Gawlikowski, Cedrique Rovile Njieutcheu Tassi, Mohsin Ali, Jongseok Lee, Matthias Humt, Jianxiang Feng, Anna Kruspe, Rudolph Triebel, Peter Jung, Ribana Roscher, Muhammad Shahzad, Wen Yang (+2 others)
2022 arXiv   pre-print
The modeling of these uncertainties based on deterministic neural networks, Bayesian neural networks, ensemble of neural networks, and test-time data augmentation approaches is introduced and different  ...  For a practical application, we discuss different measures of uncertainty, approaches for the calibration of neural networks and give an overview of existing baselines and implementations.  ...  A) single deterministic model, B) Bayesian neural network, B) ensemble approach, and D) test-time data augmentation.  ... 
arXiv:2107.03342v3 fatcat:cex5j3xq5fdijjdtdbt2ixralm

A King's Ransom for Encryption: Ransomware Classification using Augmented One-Shot Learning and Bayesian Approximation

Amir Atapour-Abarghouei, Stephen Bonner, Andrew Stephen McGough
2019 2019 IEEE International Conference on Big Data (Big Data)  
learning.  ...  Instead of creating a large training dataset of ransomware screenshots, we simulate screenshot capture conditions via carefully-designed data augmentation techniques, enabling simple and efficient one-shot  ...  [20] directly leverage data augmentation for oneshot learning.  ... 
doi:10.1109/bigdata47090.2019.9005540 dblp:conf/bigdataconf/AbarghoueiBM19 fatcat:vqaunkyg7vcw5m5sy6cxgf3jga

Research Trends and Applications of Data Augmentation Algorithms [article]

Joao Fonseca, Fernando Bacao
2022 arXiv   pre-print
In the Machine Learning research community, there is a consensus regarding the relationship between model complexity and the required amount of data and computation power.  ...  Because of this, data augmentation techniques became a popular research topic in recent years.  ...  A survey on Image Data 2019 590 Augmentation for Deep Learning Salamon J., Bello J.P.  ... 
arXiv:2207.08817v2 fatcat:d6el7fh4yned7cvoc2la2k6gpa

Hybrid Deep Learning Model using SPCAGAN Augmentation for Insider Threat Analysis [article]

R G Gayathri, Atul Sajjanhar, Yong Xiang
2022 arXiv   pre-print
Furthermore, we introduce a deep learning-based hybrid model for insider threat analysis.  ...  Results demonstrate that our proposed approach has a lower error, is more accurate, and generates substantially superior synthetic insider threat data than previous models.  ...  Bayesian approaches are linked to stochastic learning algorithms in deep learning, and this connection is used to approximate the posterior in complex models.  ... 
arXiv:2203.02855v1 fatcat:naeat2lz4jg65bp2gaw2tzp25q

Last Layer Marginal Likelihood for Invariance Learning [article]

Pola Schwöbel, Martin Jørgensen, Sebastian W. Ober, Mark van der Wilk
2022 arXiv   pre-print
The Bayesian paradigm for model selection provides a path towards end-to-end learning of invariances using only the training data, by optimising the marginal likelihood.  ...  approach might be employed for learning invariances.  ...  Acknowledgements MJ is supported by a research grant from the Carlsberg Foundation (CF20-0370). SWO acknowledges the support of the Gates Cambridge Trust for his doctoral studies.  ... 
arXiv:2106.07512v2 fatcat:5ojyoa62kra6jlw4hoaqnwqema

Transfer Learning-Based Deep Learning Models for Screening Covid-19 Infection from Chest CT Images

Dr. S. Malliga, Dr. S. V. Kogilavani, R. Deepti, S. Gowtham Krishnan, G. J. Adhithiya
2022 International Journal of Communications  
In addition, we look into the transfer learning of deep convolutional neural networks like VGG16, inceptionV3, and Xception for detecting infection in CT scans.To find the best values for hyper-parameters  ...  The study comprises of comparing and analysing the employed pre-trained CNN models. According to the data, all trained models are more than 93 percent correct.  ...  As we have a limited number of images, we have applied image augmentation techniques. When we don't have enough data to train deep learning models, image augmentation is an effective technique.  ... 
doi:10.46300/9107.2022.16.7 fatcat:w6medupp4rdmlnaii35eogzyyi

Accuracy of machine learning for differentiation between optic neuropathies and pseudopapilledema

Jin Mo Ahn, Sangsoo Kim, Kwang-Sung Ahn, Sung-Hoon Cho, Ungsoo S. Kim
2019 BMC Ophthalmology  
We compared four machine learning classifiers (our model, GoogleNet Inception v3, 19-layer Very Deep Convolution Network from Visual Geometry group (VGG), and 50-layer Deep Residual Learning (ResNet)).  ...  This study is to evaluate the accuracy of machine learning for differentiation between optic neuropathies, pseudopapilledema (PPE) and normals.  ...  Figure 1c shows the shows augmentation process. Training model We have constructed a convolutional neural network, using Google's Tensorflow deep learning framework as backend [11] .  ... 
doi:10.1186/s12886-019-1184-0 fatcat:i67g355wvbbt7fffcqej7ifw34

COVID-19 ResNet: Residual neural network for COVID-19 classification with bayesian data augmentation

Maria Baldeon calisto, Javier Sebastián Balseca Zurita, Martin Alejandro Cruz Patiño
2021 Avances en Ciencias e Ingenierías  
Moreover, to increase the accuracy of the model and overcome the data scarcity of COVID-19 images, a personalized data augmentation strategy using a three-step Bayesian hyperparameter optimization approach  ...  Furthermore, we also provide an insight into which data augmentation operations are successful in increasing a CNNs performance when doing medical image classification with COVID-19 CXR.  ...  Specifically, a three-step Bayesian hyperparameter optimization approach is applied to optimize nine hyperparameters that define the magnitude of the data augmentation operations and the model.  ... 
doi:10.18272/aci.v13i2.2288 fatcat:exw3wdhjcbgahgjwmmsddsfije

Data Augmentation for Bayesian Deep Learning

Yuexi Wang, Nicholas Polson, Vadim O. Sokolov
2022 Bayesian Analysis  
We use the theory of scale mixtures of normals to derive data augmentation strategies for deep learning.  ...  Data augmentation techniques are a natural approach to provide uncertainty quantification and to incorporate stochastic Monte Carlo search into stochastic gradient descent (SGD) methods.  ...  In this paper, following the spirit of hierarchical Bayesian modeling, we develop data augmentation strategies for deep learning with a complete data likelihood function equivalent to weighted least squares  ... 
doi:10.1214/22-ba1331 fatcat:tiydlbolnzfgxhydumwurqsm4u

Improving the Performance of Deep Learning for Wireless Localization [article]

Ramdoot Pydipaty, Johnu George, Krishna Selvaraju, Amit Saha
2020 arXiv   pre-print
Second, we show how to augment a typically small labelled data set using the unlabelled data set. We observed improved performance in DL by applying the two techniques.  ...  First, we apply automatic hyperparameter optimization to a deep neural network (DNN) system for indoor wireless localization, which makes the system easy to port to new wireless environments.  ...  Second, we present two augmentation techniques to augment the labelled data set so that more data is made available for training the deep learning models, thus improving their performance.  ... 
arXiv:2006.08925v1 fatcat:fujktne6xzcqzkwksqg6f5kyrq
« Previous Showing results 1 — 15 out of 22,222 results