scispace - formally typeset
Search or ask a question
Author

Cheng Pang

Bio: Cheng Pang is an academic researcher from Guilin University of Electronic Technology. The author has contributed to research in topics: Convolutional neural network & Upsampling. The author has an hindex of 3, co-authored 11 publications receiving 105 citations.

Papers
More filters
Journal ArticleDOI
TL;DR: This article proposes a dense lightweight network, called MADNet, for stronger multiscale feature expression and feature correlation learning, and presents a dual residual-path block (DRPB) that utilizes the hierarchical features from original low-resolution images.
Abstract: Recently, deep convolutional neural networks (CNNs) have been successfully applied to the single-image super-resolution (SISR) task with great improvement in terms of both peak signal-to-noise ratio (PSNR) and structural similarity (SSIM). However, most of the existing CNN-based SR models require high computing power, which considerably limits their real-world applications. In addition, most CNN-based methods rarely explore the intermediate features that are helpful for final image recovery. To address these issues, in this article, we propose a dense lightweight network, called MADNet, for stronger multiscale feature expression and feature correlation learning. Specifically, a residual multiscale module with an attention mechanism (RMAM) is developed to enhance the informative multiscale feature representation ability. Furthermore, we present a dual residual-path block (DRPB) that utilizes the hierarchical features from original low-resolution images. To take advantage of the multilevel features, dense connections are employed among blocks. The comparative results demonstrate the superior performance of our MADNet model while employing considerably fewer multiadds and parameters.

190 citations

Journal ArticleDOI
TL;DR: A cascading residual network (CRN) that contains several locally sharing groups (LSGs) that not only promotes the propagation of features and the gradient but also eases the model training is proposed, which outperforms most of the advanced methods while still retaining a reasonable number of parameters.
Abstract: Deep convolutional neural networks (CNNs) have contributed to the significant progress of the single-image super-resolution (SISR) field. However, the majority of existing CNN-based models maintain high performance with massive parameters and exceedingly deeper structures. Moreover, several algorithms essentially have underused the low-level features, thus causing relatively low performance. In this article, we address these problems by exploring two strategies based on novel local wider residual blocks (LWRBs) to effectively extract the image features for SISR. We propose a cascading residual network (CRN) that contains several locally sharing groups (LSGs), in which the cascading mechanism not only promotes the propagation of features and the gradient but also eases the model training. Besides, we present another enhanced residual network (ERN) for image resolution enhancement. ERN employs a dual global pathway structure that incorporates nonlocal operations to catch long-distance spatial features from the the original low-resolution (LR) input. To obtain the feature representation of the input at different scales, we further introduce a multiscale block (MSB) to directly detect low-level features from the LR image. The experimental results on four benchmark datasets have demonstrated that our models outperform most of the advanced methods while still retaining a reasonable number of parameters.

88 citations

Journal ArticleDOI
TL;DR: A novel deep residual convolutional neural network (DRCNN) for image denoising with skip connections that reduces the path length of gradient transfer, making the gradient transfer in a short path and alleviating the vanishing-gradient problem.
Abstract: Recently, convolutional neural network (CNN)-based methods have achieved impressive performance on image denoising. Notably, CNN with deeper and thinner structures is more flexible to extract the image details. However, direct stacking some existing networks is difficult to achieve satisfactory denoising performance. In this paper, we propose a novel deep residual convolutional neural network (DRCNN) for image denoising. The main structure of DRCNN is the residual block that consists of two convolutional layers, and there are skip connections between these two convolutional layers without the batch normalization operation. The skip connection not only directly transfers the input image information to the hidden layer but also reduces the path length of gradient transfer, making the gradient transfer in a short path and alleviating the vanishing-gradient problem. DRCNN is compared with several state-of-the-art algorithms, and the experimental results demonstrated its denoising effectiveness.

25 citations

Journal ArticleDOI
TL;DR: A novel Bilinear Pyramid Network (BPN) for flower categorization is presented, where features from a convolutional layer are resized and multiplied with that from the former layer, which alternates multiple times to generates prediction vectors using the features from distinct layers.
Abstract: It is a challenging task to distinguish between numerous species of flowers due to their visually similarities and variations of the pose and structure. Thanks to properly modeling of the local feature interactions, bilinear CNN has succeeded in classifying of many non-rigid fine-grained species including flowers. However, bilinear CNN only computes the feature in a straightforward way without exploring the interactions between features from multiple layers in the network. In this paper, we present a novel Bilinear Pyramid Network (BPN) for flower categorization. Instead of passing through the network and directly feeding the final classifier, features from a convolutional layer are resized and multiplied with that from the former layer, which alternates multiple times to generates prediction vectors using the features from distinct layers. These features encoded from the feature pyramid spontaneously carry multi-level semantic cues, which yields stronger discriminative powers than single-layer features. Experiments show that the proposed network obtains superior classification results on the challenging dataset of flowers.

7 citations

Journal ArticleDOI
TL;DR: A novel multi-scale rain removal model that adapts a two-branch squeeze-and-excitation residual network architecture that learns the basic structure and texture details of the corresponding clean image to effectively remove rain streaks from an image to restore its structural information and details.

2 citations


Cited by
More filters
Journal ArticleDOI
TL;DR: This article proposes a dense lightweight network, called MADNet, for stronger multiscale feature expression and feature correlation learning, and presents a dual residual-path block (DRPB) that utilizes the hierarchical features from original low-resolution images.
Abstract: Recently, deep convolutional neural networks (CNNs) have been successfully applied to the single-image super-resolution (SISR) task with great improvement in terms of both peak signal-to-noise ratio (PSNR) and structural similarity (SSIM). However, most of the existing CNN-based SR models require high computing power, which considerably limits their real-world applications. In addition, most CNN-based methods rarely explore the intermediate features that are helpful for final image recovery. To address these issues, in this article, we propose a dense lightweight network, called MADNet, for stronger multiscale feature expression and feature correlation learning. Specifically, a residual multiscale module with an attention mechanism (RMAM) is developed to enhance the informative multiscale feature representation ability. Furthermore, we present a dual residual-path block (DRPB) that utilizes the hierarchical features from original low-resolution images. To take advantage of the multilevel features, dense connections are employed among blocks. The comparative results demonstrate the superior performance of our MADNet model while employing considerably fewer multiadds and parameters.

190 citations

Journal ArticleDOI
TL;DR: The definition of fractional calculus is introduced into a 3D multi-attribute chaotic system in this research, and a novel chaotic system without equilibrium points is constructed, in which the nonlinear function term in FMACS is replaced with a rare non linear function e x .
Abstract: The definition of fractional calculus is introduced into a 3D multi-attribute chaotic system in this paper. The fractional multi-attribute chaotic system (FMACS) numerical solution is obtained based on the Adomian decomposition method (ADM). The balance points and dynamical behaviors of self-excited and hidden attractors in FMACS are compared and analyzed through the Lyapunov spectrum, bifurcation model, and complexity. It is worth noting that some hidden coexistence attractors with different shapes are affected by the order. Besides, a novel chaotic system without equilibrium points is constructed, in which the nonlinear function term in FMACS is replaced with a rare nonlinear function e x . Meanwhile, its degradation phenomenon and state transition phenomenon are analyzed in detail. Finally, the digital circuit of the system is realized on the DSP board. The research result shows that FMACS has richer dynamical behaviors and higher complexity. This research provides a theoretical basis and guidance for the application of fractional chaotic systems.

94 citations

Journal ArticleDOI
TL;DR: This method uses a pixel intensity center regionalization strategy to perform centralization of the image histogram on the overall image and dual-image multi-scale fusion to integrate the contrast, saliency and exposure weight maps of the color corrected and contrast enhanced images.
Abstract: Underwater images suffer from color cast and low visibility caused by the medium scattering and absorption, which will reduce the use of valuable information from the image. In this paper, we propose a novel method which includes four stages of pixel intensity center regionalization, global equalization of histogram, local equalization of histogram and multi-scale fusion. Additionally, this method uses a pixel intensity center regionalization strategy to perform centralization of the image histogram on the overall image. Global equalization of histogram is employed to correct color of the image according to the characteristics of each channel. Local equalization of dual-interval histogram based on average of peak and mean values is used to improve contrast of the image according to the characteristics of each channel. Dual-image multi-scale fusion to integrate the contrast, saliency and exposure weight maps of the color corrected and contrast enhanced images. Experiments on variety types of degraded underwater images show that the proposed method produces better output results in both qualitative and quantitative analysis, thus, the proposed method outperforms other state-of-the-art methods.

57 citations

Journal ArticleDOI
TL;DR: This work designs a dual-channel network through 2D and 3D convolution to jointly exploit the information from both single band and adjacent bands, which is different from previous works.
Abstract: Deep learning-based hyperspectral image superresolution methods have achieved great success recently. However, most methods utilize 2D or 3D convolution to explore features, and rarely combine the two types of convolution to design networks. Moreover, when the model only contains 3D convolution, almost all the methods take all the bands of hyperspectral image as input to analyze, which requires more memory footprint. To address these issues, we explore a new structure for hyperspectral image superresolution using spectrum and feature context. Inspired by the high similarity among adjacent bands, we design a dual-channel network through 2D and 3D convolution to jointly exploit the information from both single band and adjacent bands, which is different from previous works. Under the connection of depth split, it can effectively share spatial information so as to improve the learning ability of 2D spatial domain. Besides, our method introduces the features extracted from previous band, which contributes to the complementarity of information and simplifies the network structure. Through feature context fusion, it significantly enhances the performance of the algorithm. Extensive evaluations and comparisons on three public datasets demonstrate that our approach produces the state-of-the-art results over the existing approaches.

53 citations

Journal ArticleDOI
TL;DR: This work designs the dedicated fractions to compensate the lower color channels and an adaptive contrast enhancement algorithm is applied to each color channel to produce the background-stretched and foreground-stretched images, which significantly improves the contrast of the output image.

46 citations