Filters








17,965 Hits in 10.2 sec

It is all in the noise: Efficient multi-task Gaussian process inference with structured residuals

Barbara Rakitsch, Christoph Lippert, Karsten M. Borgwardt, Oliver Stegle
2013 Neural Information Processing Systems  
We propose a multi-task Gaussian process approach for modeling both the relatedness between regressors and the task correlations in the residuals, in order to more accurately identify true sharing between  ...  The resulting Gaussian model has a covariance term in form of a sum of Kronecker products, for which efficient parameter inference and out of sample prediction are feasible.  ...  KB was supported by the Alfried Krupp Prize for Young University Teachers of the Alfried Krupp von Bohlen und Halbach-Stiftung.  ... 
dblp:conf/nips/RakitschLBS13 fatcat:4d7t2y4dd5ghnc5k6f3qc25k5y

Reconstructing the Noise Manifold for Image Denoising [article]

Ioannis Marras, Grigorios G. Chrysos, Ioannis Alexiou, Gregory Slabaugh, Stefanos Zafeiriou
2020 arXiv   pre-print
The task of leveraging structures in the target space is unstable due to the complexity of patterns in natural scenes, so the presence of unnatural artifacts or over-smoothed image areas cannot be avoided  ...  To fill the gap, in this work we introduce the idea of a cGAN which explicitly leverages structure in the image noise space.  ...  Also, one of the advantages of our method is that it supports multi-camera noise reduction during one inference step.  ... 
arXiv:2002.04147v2 fatcat:ffieobeygbc3hm3sz6on7lfpnm

Inferring Objectives in Continuous Dynamic Games from Noise-Corrupted Partial State Observations [article]

Lasse Peters, David Fridovich-Keil, Vicenç Rubies-Royo, Claire J. Tomlin, Cyrill Stachniss
2021 arXiv   pre-print
Thus, it is suitable for downstream trajectory forecasting tasks. We demonstrate our method in several simulated traffic scenarios.  ...  In this paper, we propose a method for inferring parametric objective models of multiple agents based on observed interactions.  ...  them with additive white Gaussian noise as described in (9) .  ... 
arXiv:2106.03611v3 fatcat:sbyznmeze5bybjmloyip5r3fli

Solving Inverse Problems with a Flow-based Noise Model [article]

Jay Whang, Qi Lei, Alexandros G. Dimakis
2021 arXiv   pre-print
We empirically validate the efficacy of our method on various inverse problems, including compressed sensing with quantized measurements and denoising with highly structured noise patterns.  ...  This formulation allows us to use noise models with arbitrary dependencies as well as non-linear forward operators.  ...  MAP Formulation When the likelihood under the prior p G (x) can be computed efficiently (e.g. when it is a flow model), we can pose the inverse problem as a MAP estimation task.  ... 
arXiv:2003.08089v3 fatcat:cgn25mmbxrcbhjohv63j2z5ozm

Image De-noising with Machine Learning: A Review

Rini Smita Thakur, Shubhojeet Chatterjee, Ram Narayan Yadav, Lalita Gupta
2021 IEEE Access  
Machine learning is an important tool in the image-de-noising workflow in terms of its robustness, accuracy, and time requirement.  ...  The best de-noising results for different noise type is discussed along with future prospects. Among various Gaussian noise de-noisers, GCBD, BRDNet and PReLU network prove to be promising.  ...  If an image is corrupted with white noise, it implies that all the pixels are uncorrelated with each other.  ... 
doi:10.1109/access.2021.3092425 fatcat:xirq6soukzchvaeiugcpgxnlqi

Hybrid Deep Learning Framework for Reduction of Mixed Noise via Low Rank Noise Estimation

Dai-Gyoung Kim, Yasir Ali, Muhammad Asif Farooq, Asif Mushtaq, Muhammad Ahmad Abdul Rehman, Zahid Hussain Shamsi
2022 IEEE Access  
Secondly, convolutional neural network (CNN) is applied to the pre-processed image based on the noise statistics inferred in the first step.  ...  As a result of this step, we obtain a pre-processed image with residual noise statistics.  ...  That is, it would be enquired whether the noise statistics of the residual noise in x follows the Gaussian distribution or not.  ... 
doi:10.1109/access.2022.3170490 fatcat:3advsej5hjbvfjm5frxqxmoghm

Multiple Output Regression with Latent Noise [article]

Jussi Gillberg, Pekka Marttinen, Matti Pirinen, Antti J. Kangas, Pasi Soininen, Mehreen Ali, Aki S. Havulinna, Marjo-Riitta Marjo-Riitta Järvelin, Mika Ala-Korpela, Samuel Kaski
2016 arXiv   pre-print
Therefore, (1) explaining away the structured noise in multiple-output regression is of paramount importance.  ...  In high-dimensional data, structured noise caused by observed and unobserved factors affecting multiple target variables simultaneously, imposes a serious challenge for modeling, by masking the often weak  ...  multi-task learning ('L2/L1 MTL'), kernel regression with linear and Gaussian kernels combined with a process for removing confounding factors ('KRR with linear kernel + PEER', 'KRR with Gaussian kernel  ... 
arXiv:1410.7365v2 fatcat:bn4gvhiiifgbtchz4w5kgtnelq

Improved Noise and Attack Robustness for Semantic Segmentation by Using Multi-Task Training with Self-Supervised Depth Estimation [article]

Marvin Klingner, Andreas Bär, Tim Fingscheidt
2020 arXiv   pre-print
Our evaluation exhibits a particular novelty in that it allows to mutually compare the effect of input noises and adversarial attacks on the robustness of the semantic segmentation.  ...  We show the effectiveness of our method on the Cityscapes dataset, where our multi-task training approach consistently outperforms the single-task semantic segmentation baseline in terms of both robustness  ...  , post-processing with conditional random fields (CRFs) [14, 65] and label relax-ation [75] , and multi-scale inference [14, 73] , for further performance improvements.  ... 
arXiv:2004.11072v1 fatcat:wecxnyuhpjcgxf4igvq6uqarre

Noise Robust Online Inference for Linear Dynamic Systems [article]

Saikat Saha
2015 arXiv   pre-print
We revisit the Bayesian online inference problems for the linear dynamic systems (LDS) under non- Gaussian environment.  ...  Therefore, any inference engine should not only be robust to noise outlier, but also be adaptive to potentially unknown and time varying noise parameters; yet it should be scalable and easy to implement  ...  outlier or structural break in the process model.  ... 
arXiv:1504.05723v1 fatcat:lsejlbxi4vht5plh2j542wzjzq

Facial Attribute Capsules for Noise Face Super Resolution [article]

Jingwei Xin, Nannan Wang, Xinrui Jiang, Jie Li, Xinbo Gao, Zhifeng Li
2020 arXiv   pre-print
In the SR processing, we first generated a group of FACs from the input LR face, and then reconstructed the HR face from this group of FACs.  ...  Their performance degrades drastically when applied to real-world scenarios where the input image is always contaminated by noise.  ...  The second one is to downsample with scaling factor 8, and then add Gaussian noise with noise level 10 [37] (denote as BicN for short), where the noise level n means a standard deviation n in a pixel intensity  ... 
arXiv:2002.06518v1 fatcat:zs6355bui5aqpfric55kmv5rry

Segregating event streams and noise with a Markov renewal process model [article]

Dan Stowell, Mark D. Plumbley
2012 arXiv   pre-print
We describe an inference task in which a set of timestamped event observations must be clustered into an unknown number of temporal sequences with independent and varying rates of observations.  ...  Various existing approaches to multi-object tracking assume a fixed number of sources and/or a fixed observation rate; we develop an approach to inferring structure in timestamped data produced by a mixture  ...  Acknowledgments (Acknowledgments to be added in final version.)  ... 
arXiv:1211.2972v1 fatcat:zbcitmsa5bcblptihnzvhzuhda

Lightweight Image Restoration Network for Strong Noise Removal in Nuclear Radiation Scenes

Xin Sun, Hongwei Luo, Guihua Liu, Chunmei Chen, Feng Xu
2021 Sensors  
In order to remove the strong noise with complex shapes and high density in nuclear radiation scenes, a lightweight network composed of a Noise Learning Unit (NLU) and Texture Learning Unit (TLU) was designed  ...  its high denoising efficiency and rich texture retention.  ...  Acknowledgments: The authors are grateful for the experimental platform and resources provided by the Sichuan Province Key Laboratory of Special Environmental Robotics.  ... 
doi:10.3390/s21051810 pmid:33807719 fatcat:y73v43mpqjen3gz4cvp7sprub4

Noise-Enhanced Information Systems

Hao Chen, Lav R. Varshney, Pramod K. Varshney
2014 Proceedings of the IEEE  
This is an author-produced, peer-reviewed version of this article. The final, definitive version of this document can be found online at Proceedings of IEEE, published by IEEE.  ...  ACKNOWLEDGMENT The authors would like to thank Dr. Willard Larkin for his patience and valuable advice during the preparation of this  ...  Alternatively, a random Gaussian noise image with zero mean and σ 2 n variance is applied in the single noise framework.  ... 
doi:10.1109/jproc.2014.2341554 fatcat:rx6l6dy2crgbfohvxmdf4achmy

Model Inconsistent but Correlated Noise: Multi-view Subspace Learning with Regularized Mixture of Gaussians [article]

Hongwei Yong, Deyu Meng, Jinxing Li, Wangmeng Zuo, Lei Zhang
2018 arXiv   pre-print
To enhance the robustness of model, the complexity, non-consistency and similarity of noise in multi-view data should be fully taken into consideration.  ...  Most current MSL methods only assume a simple Gaussian or Laplacian distribution for the noise while neglect the complex noise configurations in each view and noise correlations among different views of  ...  For MSL-RMoG, the number of MoG components is set as 3 in all cases, except 2 in face experiments with Gaussian noise.  ... 
arXiv:1811.02934v1 fatcat:f5vx2uxrvnclnly2ehm37n6yxi

From data to noise to data: mixing physics across temperatures with generative artificial intelligence [article]

Yihang Wang, Lukas Herron, Pratyush Tiwary
2022 arXiv   pre-print
The results here are demonstrated for a chirally symmetric peptide and single-strand ribonucleic acid undergoing conformational transitions in all-atom water.  ...  Specifically, we work with denoising diffusion probabilistic models, and show how these models in combination with replica exchange molecular dynamics achieve superior sampling of the biomolecular energy  ...  CHE-2044165 and used XSEDE Bridges through allocation TG-CHE180053, which is supported by National Science Foundation grant number ACI-1548562.  ... 
arXiv:2107.07369v2 fatcat:mzmfpy7ia5dgzkawzlw3lb3344
« Previous Showing results 1 — 15 out of 17,965 results