Training an Active Random Field for Real-Time Image Denoising
The DOI matches its Crossref record; 49 of the 49 references it cites were found in a registry or an index.
Abstract
Many computer vision problems can be formulated in a Bayesian framework based on Markov random fields (MRF) or conditional random fields (CRF). Generally, the MRF/CRF model is learned independently of the inference algorithm that is used to obtain the final result. In this paper, we observe considerable gains in speed and accuracy by training the MRF/CRF model together with a fast and suboptimal inference algorithm. An active random field (ARF) is defined as a combination of a MRF/CRF based model and a fast inference algorithm for the MRF/CRF model. This combination is trained through an optimization of a loss function and a training set consisting of pairs of input images and desired outputs. We apply the ARF concept to image denoising, using the Fields of Experts MRF together with a 1-4 iteration gradient descent algorithm for inference. Experimental validation on unseen data shows that the ARF approach obtains an improved benchmark performance as well as a 1000-3000 times speedup compared to the Fields of Experts MRF. Using the ARF approach, image denoising can be performed in real-time, at 8fps on a single CPU for a 256 x 256 image sequence, with close to state-of-the-art accuracy.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Citations
Cites
Image quality assessment: from error visibility to structural similarity
Wang, Z., Bovik, A. C., Sheikh, H. R., & Simoncelli, E. P. (2004) · IEEE Transactions on Image Processing · doi:10.1109/tip.2003.819861 · Cited by 57,892 · cited 3×
“Other more appropriate loss functions could be used instead of the PSNR, for example the Structural Similarity Index (SSIM) [40].”
MethodGenetic algorithms in search, optimization, and machine learning
(1989) · Choice Reviews Online · doi:10.5860/choice.27-0936 · Cited by 49,010 · cited 1×
“Other optimization methods such as genetic algorithms [14] or simulated annealing [16] could be more appropriate for avoiding local optima and are subject to further investigation.”
Optimization by Simulated Annealing
Kirkpatrick, S., Gelatt, C. D., & Vecchi, M. P. (1983) · Science · doi:10.1126/science.220.4598.671 · Cited by 45,134 · cited 1×
“Other optimization methods such as genetic algorithms [14] or simulated annealing [16] could be more appropriate for avoiding local optima and are subject to further investigation.”
A Simplex Method for Function Minimization
Nelder, J. A., & Mead, R. (1965) · The Computer Journal · doi:10.1093/comjnl/7.4.308 · Cited by 28,905 · cited 1×
“We also experimented with gradient ascent, conjugate gradient and the simplex method [24].”
MethodGenetic Algorithms in Search, Optimization and Machine Learning
Goldberg, D. E., & Robson, D. (1988) · Cited by 18,391
Conditional Random Fields: Probabilistic Models for Segmenting and Labeling Sequence Data
Lafferty, J., McCallum, A., & Pereira, F. C. N. (2001) · ScholarlyCommons (University of Pennsylvania) · Cited by 12,935
Image denoising by sparse 3-D transform-domain collaborative filtering
Dabov, K., Foi, A., Katkovnik, V., & Egiazarian, K. (2007) · IEEE Transactions on Image Processing · doi:10.1109/tip.2007.901238 · Cited by 9,283 · cited 10×
“Non-local image denoising methods include [6] and especially 3D collaborative filtering (BM3D) [9], th e latter obtaining very good results with low computational expense.”
BackgroundMethodInfluentialOn the limited memory BFGS method for large scale optimization
Liu, D. C., & Nocedal, J. (1989) · Mathematical Programming · doi:10.1007/bf01589116 · Cited by 8,560 · cited 2×
“The MRF model was based on 3-cliques and inference was obtained by limited memory BFGS [21].”
MethodMartin, D., Fowlkes, C., Tal, D., & Malik, J. (2002) · Proceedings Eighth IEEE International Conference on Computer Vision. ICCV 2001 · doi:10.1109/iccv.2001.937655 · Cited by 8,216
A Non-Local Algorithm for Image Denoising
Buades, A., Coll, B., & Morel, J.-M. (2005) · 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05) · doi:10.1109/cvpr.2005.38 · Cited by 7,616 · cited 1×
“Nonlocal image denoising methods include [6] and especially 3-”
Method
Cited by
FFDNet: Toward a Fast and Flexible Solution for CNN-Based Image Denoising
Zhang, K., Zuo, W., & Zhang, L. (2018) · IEEE Transactions on Image Processing · doi:10.1109/tip.2018.2839891 · Cited by 2,678
“the discriminative architecture is derived from optimization algorithms such as HQS and gradient descent [17], [21]–[24].”
MethodInfluentialDeep Convolutional Neural Network for Inverse Problems in Imaging
Jin, K. H., McCann, M. T., Froustey, E., & Unser, M. (2017) · IEEE Transactions on Image Processing · doi:10.1109/tip.2017.2713099 · Cited by 2,529
“[34] described the usage of iterative gradient descent inferences for maximum a posteriori (MAP) estimation, and the unrolling concept came out in the derivation.”
BackgroundConditional Random Fields as Recurrent Neural Networks
Zheng, S., Jayasumana, S., Romera-Paredes, B., Vineet, V., Su, Z., Du, D., Huang, C., & Torr, P. H. S. (2015) · 2015 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2015.179 · Cited by 2,386
“[4] proposed a joint training of a MRF/CRF model together with an inference algorithm in their Active Random Field approach.”
MethodLearning Deep CNN Denoiser Prior for Image Restoration
Zhang, K., Zuo, W., Gu, S., & Zhang, L. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.300 · Cited by 2,116
“On the contrary, discriminative learning methods try to learn the prior parameters Θ and a compact inference through an optimization of a loss function on a training set containing degraded-clean image pairs [2, 13, 51, 55, 57].”
BackgroundChen, Y., & Pock, T. (2017) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2016.2596743 · Cited by 1,344
Plug-and-Play Image Restoration With Deep Denoiser Prior
Zhang, K., Li, Y., Zuo, W., Zhang, L., Van Gool, L., & Timofte, R. (2022) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2021.3088914 · Cited by 1,113
“The former aim to directly solve (2) with some optimization algorithms, while the latter mostly train a truncated unfolding inference through an optimization of a loss function on a training set containing N degraded-clean image pairs { ( y i , x i ) } Ni =1 [5], [6], [7], [8], [9].”
BackgroundDeep Unfolding Network for Image Super-Resolution
Zhang, K., Van Gool, L., & Timofte, R. (2020) · 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr42600.2020.00328 · Cited by 638
Image denoising using deep CNN with batch renormalization
Tian, C., Xu, Y., & Zuo, W. (2020) · Neural Networks · doi:10.1016/j.neunet.2019.08.022 · Cited by 589
“Weighted nuclear norm minimization (WNNM) [19], markov random field (MRF) [43, 1, 2], gradient methods [3, 69] and total-variation (TV) methods [49, 5] are also very popular image-denoising methods.”
BackgroundChen, K., Pu, X., Ren, Y., Qiu, H., Lin, F., & Zhang, S. (2021) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2020.3034752 · Cited by 561
Shrinkage Fields for Effective Image Restoration
Schmidt, U., & Roth, S. (2014) · 2014 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2014.349 · Cited by 543
“While the ARF model [2] (designed for real-time denoising) is more efficient (CPU only) than our CSF with the same number of stages, filters, and filter size, it exhibits inferior results: It performs worse than CSF 45 5 (Tab.”
BackgroundMethodInfluentialOn learning optimized reaction diffusion processes for effective image restoration
Chen, Y., Yu, W., & Pock, T. (2015) · 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2015.7299163 · Cited by 323
Dual Adversarial Network: Toward Real-World Noise Removal and Noise Generation
Yue, Z., Zhao, Q., Zhang, L., & Meng, D. (2020) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-58607-2_3 · Cited by 229
Wang, S., Celebi, M. E., Zhang, Y.-D., Yu, X., Lu, S., Yao, X., Zhou, Q., Miguel, M.-G., Tian, Y., Gorriz, J. M., et al. (2021) · Information Fusion · doi:10.1016/j.inffus.2021.07.001 · Cited by 241
RENOIR – A dataset for real low-light image noise reduction
Anaya, J., & Barbu, A. (2018) · Journal of Visual Communication and Image Representation · doi:10.1016/j.jvcir.2018.01.012 · Cited by 221
Self-citationGeneric Methods for Optimization-Based Modeling
Domke, J. (2012) · International Conference on Artificial Intelligence and Statistics · Cited by 434
A comprehensive survey on impulse and Gaussian denoising filters for digital images
Mafi, M., Martin, H., Cabrerizo, M., Andrian, J., Barreto, A., & Adjouadi, M. (2019) · Signal Processing · doi:10.1016/j.sigpro.2018.12.006 · Cited by 202
Munoz, D., Bagnell, J. A., & Hebert, M. (2010) · Lecture Notes in Computer Science · doi:10.1007/978-3-642-15567-3_5 · Cited by 199
“The notion of training the inference algorithm to make correct predictions is also similar to Barbu [2] for image denoising, in which a model is trained knowing that an inaccurate, but fast, inference algorithm will be used.”
MethodGaussian Conditional Random Field Network for Semantic Segmentation
Vemulapalli, R., Tuzel, O., Liu, M.-Y., & Chellappa, R. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.351 · Cited by 150
External Prior Guided Internal Prior Learning for Real-World Noisy Image Denoising
Xu, J., Zhang, L., & Zhang, D. D. (2018) · IEEE Transactions on Image Processing · doi:10.1109/tip.2018.2811546 · Cited by 153
“The representative methods include the neural network based methods [20]–[22], random fields based methods [23], [24], and reaction diffusion based methods [25].”
MethodHierarchical parsing and semantic navigation of full body CT data
Seifert, S., Barbu, A., Zhou, S. K., Liu, D., Feulner, J., Huber, M., Suehling, M., Cavallaro, A., & Comaniciu, D. (2009) · SPIE Proceedings · doi:10.1117/12.812214 · Cited by 137
“Using this inference algorithm, the DAN is an Active Random Field.(9) Training the MRF parameters is an optimization procedure based on simulated annealing to minimize the misclassification error.”
MethodSelf-citationInsights Into Analysis Operator Learning: From Patch-Based Sparse Models to Higher Order MRFs
Chen, Y., Ranftl, R., & Pock, T. (2014) · IEEE Transactions on Image Processing · doi:10.1109/tip.2014.2299065 · Cited by 131
“other analysis models, we compared its performance with existing analysis models, including typical FoE models [3], [11], [16], [30], [33], [34] and the currently published best analysis operator model - GOAL [17].”
Model-Blind Video Denoising via Frame-To-Frame Training
Ehret, T., Davy, A., Morel, J., Facciolo, G., & Arias, P. (2019) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2019.01163 · Cited by 108
“Other works proposed neural networks with an architecture obtained by unrolling optimization algorithms such as those used for MAP inference with MRFs probabilistic models [2, 39, 10, 43].”
MethodXiang, L., Yao, X., & Fang, Y. (2018) · IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing · doi:10.1109/jstars.2018.2865187 · Cited by 89
“Probabilistic graphical models, such as CRF [20], [21] or MRF [22]–[24], are commonly used to address this problem.”
MethodVemulapalli, R., Tuzel, O., & Liu, M.-Y. (2016) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2016.519 · Cited by 80
“[2] [22] [40] [4] [7] [8] [6] [3] [14] [31] [30] [5]”
MethodLearning non-local range Markov Random field for image restoration
Sun, J. Z., & Tappen, M. F. (2011) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2011.5995520 · Cited by 68
“Image models based on MRFs have been widely investigated and applied to vision problems, e.g., image denoising [21, 1, 13, 32, 26, 22, 27, 17], inpainting [20], segmentation [2], stereo [25], etc.”
BackgroundMethodInfluentialRevisiting Loss-Specific Training of Filter-Based MRFs for Image Restoration
Chen, Y., Pock, T., Ranftl, R., & Bischof, H. (2013) · Lecture notes in computer science · doi:10.1007/978-3-642-40602-7_30 · Cited by 70
“Domke [5] and Barbu [2] propose two similar approaches for the training of MRF model parameters also under the framework of bi-level optimization.”
BackgroundMethodInfluentialVideo Denoising by Combining Patch Search and CNNs
Davy, A., Ehret, T., Morel, J., Arias, P., & Facciolo, G. (2020) · Journal of Mathematical Imaging and Vision · doi:10.1007/s10851-020-00995-0 · Cited by 58
A New Color Filter Array With Optimal Properties for Noiseless and Noisy Color Image Acquisition
Condat, L. (2011) · IEEE Transactions on Image Processing · doi:10.1109/tip.2011.2114355 · Cited by 45
Jin, M., Roth, S., & Favaro, P. (2017) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2017.408 · Cited by 43
“Thus, the training of such network is the optimization of a gradient descent algorithm [2, 7].”
BackgroundMarquez, M. A., Lai, Y., Liu, X., Jiang, C. D., Zhang, S., Argüello, H., & Liang, J. (2022) · IEEE Journal of Selected Topics in Signal Processing · doi:10.1109/jstsp.2022.3172592 · Cited by 39
Separable Markov Random Field Model and Its Applications in Low Level Vision
Sun, J. Z., & Tappen, M. F. (2012) · IEEE Transactions on Image Processing · doi:10.1109/tip.2012.2208981 · Cited by 37
Structured Dictionary Learning for Image Denoising Under Mixed Gaussian and Impulse Noise
Zhu, H., & Ng, M. K. (2020) · IEEE Transactions on Image Processing · doi:10.1109/tip.2020.2992895 · Cited by 36
Hyperspectral Image Restoration Using Adaptive Anisotropy Total Variation and Nuclear Norms
Hu, T. W., Li, W., Liu, N., Tao, R., Zhang, F., & Scheunders, P. (2020) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2020.2999634 · Cited by 35
“Learning-based approaches remove noise by a mapping learned from pairs of clean and contaminated images [7]–[9].”
BackgroundA simple, fast and efficient approach to denoisaicking: Joint demosaicking and denoising
Condat, L. (2010) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2010.5652196 · Cited by 38
“There exist efficient denoising methods, including the Non-Local Means, for which fast implementations have been reported [18], [19].”
MethodTraining energy-based models for time-series imputation
Brakel, P., Stroobandt, D., & Schrauwen, B. (2013) · Journal of Machine Learning Research · doi:10.5555/2567709.2567749 · Cited by 41
“The optimizer is not necessarily run until convergence and a prediction is now given by V̂(i, j)∈Ω← opt-alg V(i, j)∈Ω E(V;θ).”
BackgroundMethodInfluentialConvolutional Neural Networks for Image Denoising and Restoration
Zuo, W., Zhang, K., & Zhang, L. (2018) · Advances in computer vision and pattern recognition · doi:10.1007/978-3-319-96029-6_4 · Cited by 44
Deep unfolding multi-scale regularizer network for image denoising
Xu, J., Yuan, M., Yan, D.-M., & Wu, T. (2023) · Computational Visual Media · doi:10.1007/s41095-022-0277-5 · Cited by 34
“Many unfolding methods have been proposed for various image restoration tasks, including image denoising [6, 51, 50], image deblurring [49, 11, 22], and image super-resolution [60].”
BackgroundA survey of denoising techniques for multi-parametric prostate MRI
Garg, G., & Juneja, M. (2018) · Multimedia Tools and Applications · doi:10.1007/s11042-018-6487-2 · Cited by 31
“Table 6 Subjective evaluation of denoising techniques on mp-MRI images Filter Subjective Score S Original Image Gaussian Noisy Image Rician Noisy Image T2w DCE DWI T2w DCE DWI T2w DCE DWI Anisotropic [32] 8 8 10 6 4 8 8 4 8 Median [20] 10 10 10 2 2 2 2 6 2 Wiener [25] 10 10 10 2 2 2 2 6 2 Gaussian [18] 10 10 8 2 2 2 2…”
BackgroundMethodInfluentialRotation Equivariant Proximal Operator for Deep Unfolding Methods in Image Restoration
Fu, J., Xie, Q., Meng, D., & Xu, Z. (2024) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2024.3383532 · Cited by 38
“Specifically, by establishing a direct correspondence between algorithm operators at each implementation step and network modules within each layer, one can rationally construct an almost “white box” network architecture with high interpretability.”
BackgroundAdversarial Gaussian Denoiser for Multiple-Level Image Denoising
Khan, A., Jin, W., Haider, A., Rahman, M., & Wang, D. (2021) · Sensors · doi:10.3390/s21092998 · Cited by 28
“Barbu [52] introduced the active Markov random-field (MRF) architecture by merging MRF with a faster testing process for image denoising.”
BackgroundColor Image Denoising via Discriminatively Learned Iterative Shrinkage
Sun, J. Z., & Xu, Z. (2015) · IEEE Transactions on Image Processing · doi:10.1109/tip.2015.2448352 · Cited by 27
Deep Unfolding Network for Image Desnowing With Snow Shape Prior
Guo, X., Wang, X., Fu, X., & Zha, Z.-J. (2025) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2025.3526647 · Cited by 32
“For denoising, the authors employed a combination of Markov random fields, conditional random fields, and a fast inference algorithm, all optimized using the gradient descent method [34].”
Combined ripplet and total variation image denoising methods using twin support vector machines
Shahdoosti, H. R., & Hazavei, S. M. (2017) · Multimedia Tools and Applications · doi:10.1007/s11042-017-4618-9 · Cited by 22
“The main challenge of this family is that finding the global minimum for the RFs energy function is usually NP hard (non-deterministic polynomial-time hard) [3].”
BackgroundDiscriminative Transfer Learning for General Image Restoration
Xiao, L., Heide, F., Heidrich, W., Schölkopf, B., & Hirsch, M. (2018) · IEEE Transactions on Image Processing · doi:10.1109/tip.2018.2831925 · Cited by 22
“We compare the proposed discriminative transfer learning (DTL) method with state-of-the-art image denoising techniques, including KSVD [11], FoE [26], BM3D [9], LSSC [21], WNNM [13], EPLL [35], opt-MRF [7], ARF [2], CSF [29] and TRD [8].”
MethodUnfolded Deep Kernel Estimation for Blind Image Super-Resolution
Zheng, H., Yong, H., & Zhang, L. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-19797-0_29 · Cited by 21
Synthesizing Camera Noise Using Generative Adversarial Networks
Henz, B., Gastal, E. S. L., & Oliveira, M. M. (2020) · IEEE Transactions on Visualization and Computer Graphics · doi:10.1109/tvcg.2020.3012120 · Cited by 20
“NOISE is a fundamental problem in graphics, image processing, and computer vision, and many image-denoising techniques have been proposed in recent years [1], [2], [3], [4], [5].”
BackgroundJuneja, M., Saini, S. K., Gupta, J., Garg, P., Thakur, N., Sharma, A., Mehta, M., & Jindal, P. (2021) · Multimedia Tools and Applications · doi:10.1007/s11042-021-11044-2 · Cited by 16
Learning-Based Image Reconstruction via Parallel Proximal Algorithm
Bostan, E., Kamilov, U. S., & Waller, L. (2018) · IEEE Signal Processing Letters · doi:10.1109/lsp.2018.2833812 · Cited by 16
“Through the learning of parametrization coefficients in a data-driven fashion, such algorithms have adapted the regularizer to the underlying properties (deterministic and/or stochastic) of the data [20–24].”
Cascaded and Recursive ConvNets (CRCNN): An effective and flexible approach for image denoising
Khowaja, S. A., Yahya, B. N., & Lee, S. (2021) · Signal Processing Image Communication · doi:10.1016/j.image.2021.116420 · Cited by 16
Robust Dehaze Algorithm for Degraded Image of CMOS Image Sensors
Qu, C., Du-yan, B., Sui, P., Ai-nong, C., & Wang, Y.-F. (2017) · Sensors · doi:10.3390/s17102175 · Cited by 16
Combining ARF and OR-PCA for Robust Background Subtraction of Noisy Videos
Javed, S., Bouwmans, T., & Jung, S. K. (2015) · Lecture notes in computer science · doi:10.1007/978-3-319-23234-8_32 · Cited by 15
“In this work, rainy video is denoised or restored by exploiting the ARF information with the same training data used in [9,10].”
MethodA Perceptually Inspired New Blind Image Denoising Method Using $L_{1}$ and Perceptual Loss
Uddin, A. F. M. S., Chung, T., & Bae, S. (2019) · IEEE Access · doi:10.1109/access.2019.2926848 · Cited by 14
“Barbu [22] proposed an Active Markov random field model by combining MRF with a faster inference process for image denoising.”
MethodJoint Training of Generic CNN-CRF Models with Stochastic Optimization
Kirillov, A., Schlesinger, D., Zheng, S., Savchynskyy, B., Torr, P. H. S., & Rother, C. (2017) · Lecture notes in computer science · doi:10.1007/978-3-319-54184-6_14 · Cited by 17
“Besides the approaches [11] and [12], there are many other works that consider the idea of backpropagation with a so-called unrolled CRF-inference scheme, such as [14,15,16,17,18,19,20].”
BackgroundImage blind denoising using dual convolutional neural network with skip connection
Wu, W., Liao, S., Lv, G., Liang, P., & Zhang, Y. (2025) · Signal Processing Image Communication · doi:10.1016/j.image.2025.117365 · Cited by 18
Gabor-Based Learnable Sparse Representation for Self-Supervised Denoising
Liu, S., Cheng, S., & Alkhalifah, T. A. (2024) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2024.3353315 · Cited by 12
“Deep unfolding methods are utilized in computer vision tasks [79, 84, 57], such as image denoising [5, 68, 19] and image Compressive Sensing (CS) [67, 83].”
MethodGreen Channel Guiding Denoising on Bayer Image
Tan, X., Lai, S., Liu, Y., & Zhang, M. (2014) · The Scientific World JOURNAL · doi:10.1155/2014/979081 · Cited by 12
Image denoising using common vector approach
Özkan, K., & Seke, E. (2015) · IET Image Processing · doi:10.1049/iet-ipr.2014.0979 · Cited by 11
Benchmarking learned algorithms for computed tomography image reconstruction tasks
Kiss, M. B., Biguri, A., Shumaylov, Z., Sherry, F., Batenburg, K. J., Schönlieb, C., & Lucka, F. F. (2025) · Applied Mathematics for Modern Challenges · doi:10.3934/ammc.2025001 · Cited by 12
Pattern Recognition on Radar Images Using Augmentation
Andriyanov, N. A., & Andriyanov, D. (2020) · Ural Symposium on Biomedical Engineering, Radioelectronics and Information Technology (USBEREIT) · doi:10.1109/usbereit48449.2020.9117669 · Cited by 11
Mukherjee, L., Sagar, M. A. K., Ouellette, J. N., Watters, J. J., & Eliceiri, K. W. (2021) · Biomedical Optics Express · doi:10.1364/boe.417108 · Cited by 10
“[41] proposed a joint training of CRF together with an inference algorithm in their Active Random Field approach.”
MethodEdge-guided CNN for Denoising Images from Portable Ultrasound Devices
Ma, Y., Yang, F., & Basu, A. (2021) · Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition · doi:10.1109/icpr48806.2021.9412758 · Cited by 11
“Model-based methods generally require repeated iterations to solve this formula, while discriminative learning methods learn the prior parameter θ through a loss function [25] [26] as below: Combining the two methods here, the regularization term [27] [28] corresponds to a denoising sub-problem.”
MethodGracheva, I., & Kopylov, A. (2017) · Communications in computer and information science · doi:10.1007/978-3-319-52920-2_24 · Cited by 7
Image Restoration Based on End-to-End Unrolled Network
Tao, X., Zhou, H., & Chen, Y. (2021) · Photonics · doi:10.3390/photonics8090376 · Cited by 7
Adaptive Context-Tree-Based Statistical Filtering for Raster Map Image Denoising
Chen, M., Xu, M., & Fränti, P. (2011) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2011.2166538 · Cited by 7
Cheng, Y., Shao, M., Wan, Y., Qiao, Y., Zuo, W., & Meng, D. (2025) · Neural Networks · doi:10.1016/j.neunet.2025.107700 · Cited by 7
Sharma, A. M., Dogra, A., Goyal, B., Saini, A., & Kukreja, V. (2025) · Computers, materials & continua/Computers, materials & continua (Print) · doi:10.32604/cmc.2025.066481 · Cited by 5
Edge and texture aware image denoising using median noise residue U-net with hand-crafted features
Soniya, S., & Sriharipriya, K. C. (2025) · PeerJ Computer Science · doi:10.7717/peerj-cs.2449 · Cited by 5
“To minimize an energy function that represents the properties of the image, for instance, optimization techniques are used in image denoising with Markov random fi elds (MRF) or conditional random fi elds (CRF) ( Barbu, 2009 ).”
Survey of Denoising, Segmentation and Classification of Pancreatic Cancer Imaging
Kaur, H., Saini, S. K., Thakur, N., & Juneja, M. (2023) · Current Medical Imaging Formerly Current Medical Imaging Reviews · doi:10.2174/1573405620666230515090523 · Cited by 6
Akal, O., & Barbu, A. (2019) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2019.8803801 · Cited by 5
“To avoid getting stuck in a shallow optimum, we followed [15] and used the result of a trained RNN with fewer iterations as initialization.”
MethodSelf-citationConvolutional Neural Network Combined with Half-Quadratic Splitting Method for Image Restoration
Li, R., Luo, L., & Zhang, Y. (2020) · Journal of Sensors · doi:10.1155/2020/8813413 · Cited by 5
“The fi nal target is the network output, and the target distance is the minimum [6 – 10], i.e., min Θ ℓ ð ̂ x , x Þ whose constraint condition is As shown above, we can conclude that the most obvious di ff erence between the model-based optimization approach and the discriminative learning approach…”
BackgroundDeep Unfolding of the Half-Quadratic Splitting Algorithm for ISAR Image Super-resolution
Yang, Z., Xia, J., Liu, T., Zhi, S., & Liu, Z. (2021) · CIE International Conference on Radar (RADAR) · doi:10.1109/radar53847.2021.10028343 · Cited by 5
“The deep unfolding methods in image processing have achieved certain successes, such as denoising [22], [23], deblurring [24]–[27] and demosaicking [28].”
BackgroundModel Meets Deep Learning in Image Inverse Problems
Wang, N., & Sun, J. Z. (2020) · CSIAM Transactions on Applied Mathematics · doi:10.4208/csiam-am.2020-0016 · Cited by 4
“Compared with the classical MRF model (such as FOE [22], ARF [91]), NLR-MRF has better denoising results.”
MethodPerceptual adversarial non-residual learning for blind image denoising
Khan, A., Jin, W., & Naqvi, R. A. (2022) · Soft Computing · doi:10.1007/s00500-022-06853-y · Cited by 4
“Barbu (2009) presented the active Markov random-field (MRF) model by combining the MRF with a quicker testing method for image denoising task.”
MethodCombining Fields of Experts (FoE) and K-SVD methods in pursuing natural image priors
Jiang, F., Chen, Z., Nazir, A., Shi, W., Lim, W. X., Liu, S., & Rho, S. (2021) · Journal of Visual Communication and Image Representation · doi:10.1016/j.jvcir.2021.103142 · Cited by 4
Reified Context ModelsPreprint
Steinhardt, J., & Liang, P. (2015) · arXiv (Cornell University) · doi:10.48550/arxiv.1502.06665 · Cited by 4
“, 2015) or by propagating the approximations through the gradient updates (Barbu, 2009; Domke, 2011; Stoyanov et al., 2011; Steinhardt & Liang, 2015).”
BackgroundDeep plug-and-play and deep unfolding methods for image restoration
Zhang, K., & Timofte, R. (2022) · Elsevier eBooks · doi:10.1016/b978-0-12-822109-9.00023-0 · Cited by 4
Fast 3D Liver Segmentation Using a Trained Deep Chan-Vese Model
Akal, O., & Barbu, A. (2022) · Electronics · doi:10.3390/electronics11203323 · Cited by 3
“To avoid getting stuck in a shallow optimum, we used the result of a trained RNN with fewer iterations as initialization, similar to [34].”
MethodSelf-citationA Markov random field based method for removing invalid unwrapping phase points in 3D reconstruction
Gao, J., Wu, F., Cheng, C., Wu, C., & Zhou, Y. (2023) · IET Image Processing · doi:10.1049/ipr2.12879 · Cited by 3
Hierarchical Object Parsing from Structured Noisy Point Clouds
Barbu, A. (2012) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2012.262 · Cited by 3
“The contributions of the paper are the following: First, it presents a hierarchical generative model that represents the object shape as a MRF-based deformation from a PCA backbone, obtaining a more accurate boundary delineation than with the PCA model alone.”
BackgroundInfluentialSelf-citationAutism Detection using r-fMRI: Subspace Approximation and CNN Based Approach
Gupta, D. S. (2020) · International Journal of Advanced Trends in Computer Science and Engineering · doi:10.30534/ijatcse/2020/20912020 · Cited by 3
Toward a Binarization Framework resolving the Maghrebian Font Database challenges
Omar, I. O., Haboubi, S., & Benzarti, F. (2020) · International Multi-Conference on Systems, Signals & Devices (SSD) · doi:10.1109/ssd49366.2020.9364086 · Cited by 2
“Gibbs distribution proved to obtain the best accuracy within image enhancement approaches [1][7][21].”
BackgroundHigh-Order MRF Prior Based Bayesian Deblurring
Zhao, B., -, W. Z., Liu, J., & Ding, H. W. (2011) · International Journal of Digital Content Technology and its Applications · doi:10.4156/jdcta.vol5.issue12.47 · Cited by 2
“It has been widely exploited across low-level vision problems, e.g., image deblurring [10], image denoising [11, 12], super resolution [13], image inpainting [11, 14] etc. Boykov et al. [15] propose pairwise MRFs to model the statistical dependency between the neighboring pixels.”
BackgroundAn end-to-end adaptive neural network for process-aware snapshot compressive temporal imaging
Marquez, M. A., Lai, Y., Liu, X., Jiang, C. D., Zhang, S., Argüello, H., & Liang, J. (2023) · High-Speed Biomedical Imaging and Spectroscopy · doi:10.1117/12.2650841 · Cited by 1
“Given access to rich datasets, many novel methods based on convolutional neural networks (CNNs) have been developed for SCTI’s reconstruction as well as for the coded-aperture design [6-9].”
MethodA New Improved Deep Unfolding Network for Single Image Deraining
Wang, L., Chen, Y., & Li, H. (2025) · Chinese Control and Decision Conference (CCDC) · doi:10.1109/ccdc65474.2025.11090485 · Cited by 1
“The early deep unfolding methods can be traced back to the image denoising via a compact maximum a posteriori inference based on gradient descent algorithm [26].”
Rectification of Noise, Background Illumination and Shadow Artifacts
Singhal, S., & Thakre, V. V. (2014) · Cited by 1
Jiang, F., Wang, X., & Zhao, D. (2013) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2013.6738020 · Cited by 1
“They are two complementary pursuits of natural images space.”
BackgroundBi-Directional Filter for the Removal of Lines and Cracks in Images
Awad, A. S. (2020) · Symmetry · doi:10.3390/sym12081280 · Cited by 1
Bu, S., Li, Y., Ren, W., & Liu, G. (2023) · Mathematical Biosciences & Engineering · doi:10.3934/mbe.2023871 · Cited by 2
Improved Attention-based Fusion Model Integrating DnCNN and Unet for Low-Light Image Enhancement
Peng, Y. (2024) · IEEE International Conference on Electronic Technology, Communication and Information (ICETCI) · doi:10.1109/icetci61221.2024.10594098 · Cited by 1
“Markov random fields (MRF) [25][26][27], gradient techniques [28], and total variation methods [29][30] are also commonly employed for image denoising.”
BackgroundHierarchical Object Parsing from Noisy Point Clouds
Barbu, A. (2011)
Self-citationRobust perceptual organization techniques for analysis of color images
Moreno, R. (2010) · UPCommons institutional repository (Universitat Politècnica de Catalunya) · doi:10.5821/dissertation-2117-93277
Multidimensional Object Parsing From StructuredNoise
Kohila, P., & Dhanalakshmi, S. (2014) · International Journal of Innovative Research in Science Engineering and Technology
Image Denoising Based on Non-parametric ADMM Algorithm
Ye, X., Zhang, M., Yan, Q., Fan, X., & Luo, Z. (2018) · Lecture notes in computer science · doi:10.1007/978-3-030-00767-6_30
L1 norm based double-prior learning for image denoising
Du, W., Sun, Y., & Hu, Y. (2019) · doi:10.1145/3357254.3357286
“Real-time image denoising problem is studied in [1] to get efficient denoising result.”
BackgroundDébruitage vidéo et applications
Ehret, T. (2020) · theses.fr (ABES) · doi:10.70675/df7a79fbz0d2dz45acz84fbz67c834568df2
Single image denoising based on adaptive fusion dual‐domain network
Jiang, Z., Xue, Z., Wang, J., Hu, Y., & Zheng, Q. (2023) · IET Image Processing · doi:10.1049/ipr2.12969
Blind Denoising Using Dense in Dense Network with Attention Module
Guo, J.-M., Budiono, D. F., Zeng, Y., & Chen, Z.-Y. (2025) · Proceedings - International Conference on Image Processing · doi:10.1109/icip55913.2025.11084468
Metadata
- Authors
- Barbu, A.
- Affiliations
- Florida State University (US)
- Corresponding author
- Barbu, A.
- ORCID
- Barbu, A.
- Venue
- IEEE Transactions on Image Processing
- Published
- 2009-11
- Volume
- 18
- Issue
- 11
- Pages
- 2451-2462
- ISSN
- 1057-7149 (ISSN-L)
- 1941-0042
- OpenAlex
- W2119290843
- Cited by
- 105
- References
- 49
- Citations by year
- 2026: 2 · 2025: 8 · 2024: 3 · 2023: 6 · 2022: 6 · 2021: 10 · 2020: 13 · 2019: 6 · 2018: 13 · 2017: 7 · 2016: 3 · 2015: 6 · 2014: 5 · 2013: 3 · 2012: 3
- FWCI
- 5.15
- Type
- journal article
- Language
- en
- Open access
- closed
- Locations
- IEEE Transactions on Image Processing (publishedVersion)
- PubMed (publishedVersion)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Image and Signal Denoising Methods
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Image Enhancement Techniques
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Image Processing Techniques
- Keywords
- image denoising, Markov random fields, conditional random fields, real-time image processing
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.