Dual Adversarial Network: Toward Real-World Noise Removal and Noise Generation

Yue, Zongsheng; Zhao, Qian; Zhang, Lei; Meng, Deyu

doi:10.1007/978-3-030-58607-2_3

Cited by 182 publications

(118 citation statements)

References 41 publications

Supporting

Mentioning

118

Contrasting

Order By: Relevance

“…Image enhancement is one of the essential components in image processing and image-display applications [18], [19]. Concerning deep learning-based distorted image enhancement, single image enhancement [5], [6], [7], [8], [9], [10], [11], and the multi-frame enhancement [12], [13] [7]. In RIDNet, local skip connections, short skip connections, and long skip connections are utilized to exploit low-frequency information over the feedforward.…”

Section: B Deep Learning-based Distorted Image Enhancementmentioning

confidence: 99%

“…ADNet combines dilated convolutions, standard convolutions, and an attention mechanism for real-noisy image denoising and blind denoising. In [9], both noise removal and noise generation tasks are trained in a Bayesian network that learns the joint distribution of the pairs of the clean and distorted images. The authors [10] propose a Block artifact removing convolutional neural networks (BARCNN) for JPEG image enhancement.…”

Section: B Deep Learning-based Distorted Image Enhancementmentioning

confidence: 99%

“…Images are sensed and encoded in IoT cameras before being sent to the IoT cloud. Unlike the existing image enhancement works, where enhancing the image relies on a single degraded image [5], [6], [7], [8], [9], [10], [11] or neighboring frames [12], [13] [14], Spectral Projected-Gradient for L1 (SPGL1) [15], Orthogonal Matching Pursuit (OMP) [16], and Sparsity Adaptive Matching Pursuit (SAMP) [17].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

CSIE-M: Compressive Sensing Image Enhancement Using Multiple Reconstructed Signals for Internet of Things Surveillance Systems

Pham

Yang

Zhou

2022

IEEE Trans. Ind. Inf.

View full text Add to dashboard Cite

Artificial Intelligence of Things (AIoT) has brought artificial intelligence (AI) to the cutting-edge Internet of Things (IoT). In recent years, Compressive sensing (CS), which relies on sparsity, is widely embedded and expected to bring more energy efficiency and a longer battery lifetime to IoT devices. Different from the other image compression standards, CS can get various reconstructed images by applying different reconstruction algorithms on coded data. Using this property, it is the first time to propose a deep learning-based compressive sensing image enhancement framework using multiple reconstructed signals (CSIE-M). Firstly, images are reconstructed by different CS reconstruction algorithms. Secondly, reconstructed images are assessed and sorted by a No-reference quality assessment module before being inputted to the quality enhancement module by order of quality scores. Finally, a multiple-input recurrent dense residual network is designed for exploiting and enriching the useful information from the reconstructed images. Experimental results show that CSIE-M obtains 1.88-8.07dB PSNR improvement while the state-of-the-art works achieve a 1.69-6.69 dB PSNR improvement under sampling rates from 0.125 to 0.75. On the other hand, using multiple reconstructed versions of the signal can improve 0.19-0.23 dB PSNR, and only 4% reconstructing time is increasing compared to using a reconstructed signal. Index Terms-Compressive sensing, Deep Learning approach for compressed image enhancement, multiple-to-one mapping.

show abstract

Section: B Deep Learning-based Distorted Image Enhancementmentioning

confidence: 99%

Section: B Deep Learning-based Distorted Image Enhancementmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

CSIE-M: Compressive Sensing Image Enhancement Using Multiple Reconstructed Signals for Internet of Things Surveillance Systems

Pham

Yang

Zhou

2022

IEEE Trans. Ind. Inf.

View full text Add to dashboard Cite

show abstract

“…Xu et al [19] developed a trilateral weighted sparse coding scheme for robust real-world image denoising. Yue et al [20] proposed a novel unified framework to simultaneously deal with noise removal and generation. Cai et al [21] proposed a novel deep generative model equipped with a brand new style extractor which can extract style features from ground truth values.…”

Section: Introductionmentioning

confidence: 99%

“…Moreover, nanoparticle superresolution reconstruction is more complicated than image reconstruction under natural light, which increases the computational complexity. As the parameters of the degradation model are unknown, the reconstruction method based on deep learning is worse than what can be achieved with conventional methods [19,20]. Although inpainting can provide a good visualization effect on image reconstruction, it is not suitable for revealing the microscopic properties of nanoparticles.…”

Section: Introductionmentioning

confidence: 99%

Image Reconstruction Using Variable Exponential Function Regularization for Wide-Field Polarization Modulation Imaging

Gao

Zhang

et al. 2021

IEEE Access

View full text Add to dashboard Cite

Polarization modulation imaging technology plays an important role in microscopic superresolution imaging. However, the specimen medium contains retardancy, while charge-coupled devices may provide discrete under-sampling, and the coupled wavefronts consisting of the polarization state of the light and the anisotropic distribution of the specimen can lead to vectorial phase fitting degradation. Considering that the point spread function (PSF) of the main degradation parts can be regarded as an asymmetric generalized Gaussian distribution with uncertain parameters, an adaptive image reconstruction method is proposed based on variable exponential function regularization. The proposed method concentrates on the diversity of the PSF and uses a variable exponent regularization to improve flexibility of the kernel. Moreover, it can balance image edge preservation and provide staircase artifact suppression, which reduces the over-and under-reconstruction of the microscopic images effectively. By optimizing the Split-Bregman algorithm, we create an efficient method that minimizes the iterative loss function under the premise of achieving high estimation accuracy. Compared with other methods, the experimental results reveal better effectiveness and robustness of the proposed method, with improvements of 18% in the peak signal-to-noise ratio, 21% in the structural similarity index measurement, and 337% in the mean structural similarity index measurement.

show abstract

Contrastive Monotonic Pixel-Level Modulation

Lü

Zhang

2022

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Dual Adversarial Network: Toward Real-World Noise Removal and Noise Generation

Cited by 182 publications

References 41 publications

CSIE-M: Compressive Sensing Image Enhancement Using Multiple Reconstructed Signals for Internet of Things Surveillance Systems

CSIE-M: Compressive Sensing Image Enhancement Using Multiple Reconstructed Signals for Internet of Things Surveillance Systems

Image Reconstruction Using Variable Exponential Function Regularization for Wide-Field Polarization Modulation Imaging

Contrastive Monotonic Pixel-Level Modulation

Contact Info

Product

Resources

About