Recursive Network Research Articles

Recursive Additive Complement Networks for Cell Membrane Segmentation in Histological Images.

A recursive additive complement network (RacNet) is introduced to segment cell membranes in histological images as closed lines. Segmenting cell membranes as closed lines is necessary to calculate cell areas and to estimate N/C ratio, which is useful to diagnose early hepatocellular carcinoma. The RacNet is composed of a complement network and an element-wise maximization (EWM) process and is recursively applied to the network output. The complement network complements the lacking parts of cell membranes. The network, however, has a tendency to mistakenly delete some parts of the segmented cell membranes. The EWM process eliminates this unwanted effect.Experiments carried out using unstained hepatic sections showed that the accuracy for segmenting cell membranes as closed lines was significantly improved by using the RacNet.Three imaging methods, bright-field, dark-field, and phase-contrast, were used, as unstained sections show very low contrast in the bright-field imaging commonly used in pathological diagnosis. These imaging methods are available in optical microscopes used by pathologists. Among the three methods, phase-contrast imaging showed the highest accuracy.

Structure‐aware 3D reconstruction for cable‐stayed bridges: A learning‐based method

AbstractA powerful deep learning‐based three‐dimensional (3D) reconstruction method for reconstructing structure‐aware semantic 3D models of cable‐stayed bridges is proposed herein. Typically, conventional bridge semantic 3D model reconstruction methods are not robust when low‐quality point clouds are used. Furthermore, they are suited particularly for their respective fields and less generalized for cable‐stayed bridges. Hence, a structure‐aware learning‐based cable‐stayed bridge 3D reconstruction framework is proposed. The encoder part of the network uses both multiview images and a photogrammetric point cloud as input, whereas the decoder part uses a recursive binary tree network to model a high‐level structural relation graph and low‐level 3D geometric shapes. Two actual cable‐stayed bridges are employed as examples to evaluate the proposed method. Test results demonstrate that the proposed method successfully reconstructs the bridge model with structural components and their relations. Quantitative results indicate that the predicted models achieved an average F1 score of 99.01%, a Chamfer distance of 0.0259, and a mesh‐to‐cloud distance of 1.78 m. The achieved result is similar to that obtained using the manual reconstruction approach in terms of component‐wise accuracy, and it is considerably better than that obtained using the manual approach in terms of spatial accuracy. In addition, the proposed recursive binary tree network is robust to noise and partial scans. The potential applications of the obtained 3D bridge models are discussed.

Inferring phenotypic causal structures among body weight traits via structural equation modeling in Kurdi sheep

Data collected on 2550 Kurdi lambs originated from 1505 dams and 149 sires during 1991 to 2015 in Hossein Abad Kurdi Sheep Breeding Station, located in Shirvan city, North Khorasan province, North-eastern area of Iran, were used for inferring causal relationship among the body weights at birth (BW), at weaning (WW), at six-month age (6MW), at nine-month age (9MW) and yearling age (YW). The inductive causation (IC) algorithm was employed to search for causal structure among these traits. This algorithm was applied to the posterior distribution of the residual (co)variance matrix of a standard multivariate model (SMM). The causal structure detected by the IC algorithm coupling with biological prior knowledge provides a temporal recursive causal network among the studied traits. The studied traits were analyzed under three multivariate models including SMM, fully recursive multivariate model (FRM) and IC-based multivariate model (ICM) via a Bayesian approach by 100,000 iterations, thinning interval of 10 and the first 10,000 iterations as burn-in. The three considered multivariate models (SMM, FRM and ICM) were compared using deviance information criterion (DIC) and predictive ability measures including mean square of error (MSE) and Pearson's correlation coefficient between the observed and predicted values (r(y, )) of records. In general, structural equation based models (FRM and ICM) performed better than SMM in terms of lower DIC and MSE and also higher r(y, ). Among the tested models ICM had the lowest (36678.551) and SMM had the highest (36744.107)DIC values. In each case of the traits studied, the lowest MSE and the highest r(y, ) were obtained under ICM. The causal effects of BW on WW, WW on 6MW, 6MW on 9MW and 9MW on YW were statistically significant values of 1.478, 0.737, 0.776 and 0.929 kg, respectively (99% highest posterior density intervals did not include zero).

Hierarchical dense recursive network for image super-resolution

Image super-resolution (SR) techniques with deep convolutional network (CNN) have achieved significant improvements compared to previous shallow-learning-based methods. Especially for dense connection based networks, these methods have yielded unprecedented achievements but bring the higher complexity and more parameters. To this end, this paper considers both reconstruction performance and efficiency, and advocates a novel hierarchical dense connection network (HDN) for image SR. First of all, we construct a hierarchical dense residual block (HDB) to promote the feature representation while saving the memory footprint with a hierarchical matrix structure design. In this way, it can provide additional interleaved pathways for information fusion and gradient optimization but with a shallower depth compare to the previous networks. In particular, a group of convolutional layers with small size (1 × 1) are embedded in HDB, releasing the computational burden and parameters by rescaling the feature dimensions. Furthermore, HDBs are connected to each other in a sharing manner, thereby allowing the network to fuse the features in different stages. At the final, the multi-scale features from these HDBs are integrated into global fusion module (GFM) for a global fusion and representation, and then the final profile-enriched residual map is obtained by realigning and sub-pixel upsampling the fusion maps. Extensive experimental results on benchmark datasets and really degraded images show that our model outperforms the state-of-the-art methods in terms of quantitative indicators and realistic visual effects, as well as enjoys a fast and accurate reconstruction.

Fractality and the small-world property of generalised (u, v)-flowers

• We generalise ( u, v )-flowers by introducing probabilities into u and v . • Small-world property provides different values of different fractal dimensions. • Value of fractal cluster dimension reflects the behaviour of the mean path length. • Value of fractal box dimension reflects the behaviour of the diameter of the graph. So-called ( u, v )-flowers are recursive networks which produce self-similar structures with fractality or the small-world property. This paper generalises ( u, v )-flowers by introducing probabilities into the realisations of u and v , which enables the study of intermediate states between small and non-small worlds in fractal networks. We obtain the analytical relation between the diameter of the graph L and the graph size N , L ∼ N 1 / d L , and the degree distribution with a power-law form. We show that the difference between the fractal cluster d c and the fractal box d b dimensions reflects different behaviour of the mean path length 〈 l 〉 and L. There seems to be an apparent contradiction between fractality and the small-world property. However, the small-world property can be reconciled with fractality of the graph by size-dependent fractal dimensions where d b shows a size-dependent increase with an upper limit d L . The invariance and equivalence of d c , d b and d L are maintained only when both 〈 l 〉 and L are subject to the same non-small-world behaviour. Our investigation provides useful information for interpreting empirical fractal data and basic tools for studying the various dynamics that occur in networks.

LSHR-Net: A hardware-friendly solution for high-resolution computational imaging using a mixed-weights neural network

Recent work showed neural-network based approaches to reconstructing images from compressively sensed measurements offer significant improvements in accuracy and signal compression. Such methods can dramatically boost the capability of computational imaging hardware. However, to date, there have been two major drawbacks: (1) the high-precision real-valued sensing patterns proposed in the majority of existing works can prove problematic when used with computational imaging hardware such as a digital micromirror sampling device and (2) the network structures for image reconstruction involve intensive computation, which is also not suitable for hardware deployment. To address these problems, we propose a novel hardware-friendly solution based on mixed-weights neural networks for computational imaging. In particular, learned binary-weight sensing patterns are tailored to the sampling device. Moreover, we proposed a recursive network structure for low-resolution image sampling and high-resolution reconstruction scheme. It reduces both the required number of measurements and reconstruction computation by operating convolution on small intermediate feature maps. The recursive structure further reduced the model size, making the network more computationally efficient when deployed with the hardware. Our method has been validated on benchmark datasets and achieved state of the art reconstruction accuracy. We tested our proposed network in conjunction with a proof-of-concept hardware setup.

Non-Global Shared Recursive Network for Image Super-Resolution

In Super-Resolution(SR) reconstruction, some recursive networks use shared parameters to keep lightweight. But these methods lack flexibility while use shared parameters to process different feature maps. We propose a non-global shared recursive network that consists of recursive units and channel attention modules. Firstly, we propose multi-scale channel attention module to learn the inter-channel interdependence and spatial information within the channel. The multi-scale channel attention module is composed of a multi-scale pooling layer, fully connected layers and activation functions. In order to flexibly process different feature maps in different layers, the attention module parameters are not shared, but only a small number of parameters are used. Secondly, residual dense recursive unit is used to accelerate convergence and feature reuse. In order to further reduce parameters, bottleneck layer is added to the dense connection. Based on the above mechanism, we propose a lightweight and efficient network structure. Experiments show that the proposed method has better performance than other lightweight networks in the standard dataset.

Dual-Stream Multi-Path Recursive Residual Network for JPEG Image Compression Artifacts Reduction

JPEG is the most widely used lossy image compression standard. When using JPEG with high compression ratios, visual artifacts cannot be avoided. These artifacts not only degrade the user experience but also negatively affect many low-level image processing tasks. Recently, convolutional neural network (CNN)-based compression artifact removal approaches have achieved significant success, however, at the cost of high computational complexity due to an enormous number of parameters. To address this issue, we propose a dual-stream recursive residual network (STRRN) which consists of structure and texture streams for separately reducing the specific artifacts related to high-frequency or low-frequency image components. The outputs of these streams are combined and fed into an aggregation network to further enhance the restored images. By using parameter sharing, the proposed network reduces the total number of training parameters significantly. Moreover, experiments conducted on five commonly used datasets confirm that the proposed STRRN can efficiently reduce the compression artifacts, while using up to 4.6 times less training parameters and 5 times less running time compared to the state-of-the-art approaches.

Wavelet Based Deep Recursive Pyramid Convolution Residual Network for Single Image Rain Removal

Image rain removal aims to separate the background image from the rainy image. During the past three years, the image rain removal with deep convolutional neural networks has achieved impressive performance. However, how to reach tradeoff between high de-raining performance and low model parameters is still a challenge. To address the issue, the paper is devoted to exploring a novel method based on wavelet deep recursive pyramid convolution residual network (WDRPRN), in which discrete wavelet transform is embedded to decompose the rainy image in different frequency domains, and the deep recursive pyramid convolution residual network (DRPRN) can well predict the residual coefficients between rainy image and clean image. In addition, compared with other neural networks, the DRPRN adopts recursive model that can cost fewer parameters. Plentiful of experiments on synthetic and real-world datasets show that the proposed method is significantly superior to the recent state-of-the-art algorithms.

REDN: A Recursive Encoder-Decoder Network for Edge Detection.

In this paper, we introduce REDN: A Recursive Encoder-Decoder Network with Skip-Connections for edge detection in natural images. The proposed network is a novel integration of a Recursive Neural Network with an Encoder-Decoder architecture. The recursive network enables iterative refinement of the edges using a single network model. Adding skip-connections between encoder and decoder helps the gradients reach all the layers of a network more easily and allows information related to finer details in the early stage of the encoder to be fully utilized in the decoder. Based on our extensive experiments on popular boundary detection datasets including BSDS500 [1], NYUD [2] and Pascal Context [3], REDN significantly advances the state-of-the-art on edge detection regarding standard evaluation metrics such as Optimal Dataset Scale (ODS) F-measure, Optimal Image Scale (OIS) F-measure, and Average Precision (AP).

Deep Recursive Network for Hyperspectral Image Super-Resolution

Fusion based hyperspectral image (HSI) super-resolution method, which obtains a spatially high-resolution (HR) HSI by fusing a low-resolution (LR) HSI and an HR conventional image, has been a prevalent method for HSI super-resolution. One effective fusion based method is to cast HSI super-resolution into a unified optimization problem, where handcrafted priors such as sparse prior or low rank prior are always adopted to regularize the latent HR HSI to be optimized. However, these priors show limitations in generalizing to challenging cases due to the heuristic assumption on image statistics as well as the restricted expressiveness capacity of the shallow structure. Taking advantages of the powerful expression ability of deep learning based method, a new HSI super-resolution network is proposed which implicitly incorporates a deep structure as the regularizer/prior. Specifically, we reformulate the original unified optimization problem into three sub-optimization problems, one is related with the regularizer and the others are without. Thanks to the fact that the one related with the regularizer naturally equals to a denoising problem, a recursive residual network is proposed for this sub-optimization problem. In addition, we unfold the other sub-optimization problems into network representations, with which the original unified optimization problem can be represented into a fully end-to-end network. Experimental results shows the superiority of the proposed method for HSI super-resolution on three benchmark datasets.

CNN Architecture for Surgical Image Segmentation with Recursive Structure and Flip-Based Upsampling

Laparoscopic surgery, a less invasive camera-aided surgery, is now performed commonly. However, it requires a camera assistant who holds and maneuvers a laparoscope. By controlling the laparoscope automatically using a robot, a surgeon can perform the operation without a camera assistant, which would be beneficial in areas suffering from lack of surgeons. In this paper, a prototype image segmentation architecture based on a convolutional neural network (CNN) is proposed to realize an automated laparoscope control for cholecystectomy. Since a training dataset is annotated manually by a few surgeons, its scale is limited compared to common CNN-based systems. Therefore, we built a recursive network structure, with some sub-networks which are used multiple times, to mitigate overfitting. In addition, instead of the common transposed convolution, the flip-based subpixel reconstruction is introduced into upsampling layers. Furthermore, we applied stochastic depth regularization to the recursive structure for better accuracy. Evaluation results revealed that these improvements bring better classification accuracy without increasing the number of parameters. The system shows a throughput sufficient for real-time laparoscope robot control with a single NVIDIA GeForce GTX 1080 GPU.

Intelligent Learning Model of Financial Spoken English Teaching based on BPTT algorithm and LSTM Network Model

It is of great research value and practical significance to use new technology to improve the accuracy of English speech recognition and apply the system to mobile platforms for users to use. The main content of this paper is the long-term and short-term memory, and the current decoding part is applied to the Android platform, and the performance of the program is analyzed. Neural networks converge slowly, making learning long-term memory difficult. In the experiment, the BPTT algorithm is used to analyze the problem of error elimination in traditional recursive networks. Combining BPTT algorithm in LSTM network to solve the problem of traditional error elimination and improve speech recognition rate. In addition, this paper uses a new LSTM recurrent neural network to study the implementation of LSTM network on Android platform. Finally, this paper designs a comparative experiment to analyze the efficiency of oral English recognition. The results show that the research algorithm of this paper has certain effects.

基于全卷积递归网络的弱小目标检测方法

基于全卷积递归网络的弱小目标检测方法

Der populäre Pakt. Verhandlungen der Moderne zwischen Operette und Feuilleton by Ethel Matala de Mazza

Reviewed by: Der populäre Pakt. Verhandlungen der Moderne zwischen Operette und Feuilleton by Ethel Matala de Mazza Sean Franzel Der populäre Pakt. Verhandlungen der Moderne zwischen Operette und Feuilleton. By Ethel Matala de Mazza. Frankfurt am Main: Fischer, 2018. Pp. 480. Cloth €25.00. ISBN 978-3103972344. This book is a sustained exploration of nineteenth- and early twentieth-century “small” forms across journalism, criticism, and popular theater. Ethel Matala de Mazza foregrounds the function of a range of “minor” genres in reflecting upon the democratic promise to involve mass publics in cultural and political life, what she calls the “popular pact.” A central goal of this book is thus to reconstruct the history of modernity as a discontinuous, fractured history of distinct public spheres; as Matala de Mazza proposes, small forms’ mosaic-like structure better captures such [End Page 612] historical discontinuities than grand narratives. The public spaces (Öffentlichkeiten) of modernity enable contestations and negotiations (Verhandlungen) between different representational modes and media. Tracking these negotiations entails attending to how “low,” popular forms play themselves off of “high,” serious forms, and how the self-images of mass culture remain ambiguous and capable of accommodating various political status quos. The concerns of this book draw on Matala de Mazza’s substantial comparatist work on figurations of the political body. This book is likewise an important touchstone for a growing body of recent studies of small forms (the author plays a leadership role at the Berlin-based Graduiertenkolleg on the “Literatur- und Wissensgeschichte kleiner Formen”). Der populäre Pakt is a compelling mixture of intellectual and cultural history, with a clear emphasis on the Frankfurt School and its interlocutors (Lukács, Weber, Schmitt). But rather than providing a genealogy of an intellectual school, the book is a history of specific engagements with popular culture in particular historical and medial contexts. It is also a genealogy of a set of critical forms. It is here that Siegfried Kracauer, and to a lesser extent Heine, Benjamin, and Karl Kraus, are revealed as heroes of the book (Adorno less so). Matala de Mazza deliberately does not ground the project in a larger social-theoretical framework (a familiar component of other Foucauldian, Weberian, or Luhmannian habilitation-like books), but she makes sure to highlight the productive relationships of a variety of “minor” literary/critical forms— feuilleton, tableau, physiology, etc.—to the nascent social thought of the nineteenth and twentieth centuries. Similar to Gerhart von Graevenitz’s 2014 monograph on Fontane, this book probes the relationship of nineteenth-century literary aesthetics to social theory in original ways. The structure of Der populäre Pakt is intriguing. It opens in the 1920s, cycles back to the early nineteenth century, and then winds its way back into the 1930s. Matala de Mazza stages a recursive network of shared concerns, medial formats, and ambivalent political constellations. She thereby enriches the story about how canonical twentieth-century figures engaged with the previous century and how nineteenth-century constellations continue to remain relevant in later historical periods. The first of the book’s three parts is an extended study of Kracauer, beginning with his feuilleton series in the Frankfurter Zeitung. Kracauer develops his poetology of the small form in response to the new mass publics of the 1920s. Matala de Mazza reconstructs Kracauer’s dialogues with Weber and Schmitt about the modern Verwaltungsstaat and she draws out similarities between the feuilleton format and the “Revue” as a genre of popular theater: both rely on the structuring principle of “bunte Mischung.” This first part closes with a reading of Kracauer’s book on Jacques Offenbach that he wrote in Parisian exile. Kracauer’s treatment of Offenbach’s world as an inverse image of the 1930s enables the book’s pivot back to the nineteenth century. The second part opens by surveying attempts to reconfigure popular festivity in [End Page 613] the wake of the French Revolution amidst debates about the possibilities of a new civic theater (Mercier, Diderot, and Rousseau). Along with exploring revolutionary festivals, the author presents the Parisian carnival as an interesting case study of the ambivalent “negotiations” of modern cultural politics. Carnival mixes old and new festival cultures and...

Image super-resolution via multi-path recursive convolutional network

Convolutional neural network (CNN) has recently achieved a great success for single image super-resolution (SISR). However, most deep CNN-based super-resolution models use chained stacking to build the network, which results in the fact that the relationship between layers is weak and does not make full use of hierarchical features. In this paper, a multi-path recursive convolutional network (MRCN) is designed to address these problems in SISR. By using multi-path structure to strengthen the relationship between layers, our network can effectively utilize features and extract rich high-frequency components. At the same time, we also use recursive structure to alleviate training difficulty. In addition, by introducing the operation of feature fusion into the model, our network can make full use of the features extracted from each layer in the reconstruction process and select the effective features adaptively. Extensive experiments on benchmarks datasets have shown that MRCN has a significant performance improvement against existing methods.

SCRSR: An efficient recursive convolutional neural network for fast and accurate image super-resolution

Convolutional neural networks have recently demonstrated high-quality reconstruction for single image super-resolution (SR). These CNN networks effectively recover a high-resolution (HR) image from a low-resolution (LR) image, at the cost of enormous parameters and heavy computational burden. In this work, we propose a recursive efficient deep convolutional network for fast and accurate single-image SR with only 0.28M parameters. A Split-Concatenate-Residual (SCR) block is proposed to reduce computation and parameters. With downsampling block and upsampling block, we significantly reduce computational complexity and enlarge the size of the receptive field. Specifically, two-level recursive learning is proposed which can improve accuracy by increasing depth without adding any weight parameters. We also employ local, semi-global and global residual techniques to train our very deep network steadily and improve its performance. Extensive experiments indicate that our proposed method Split-Concatenate-Residual Super Resolution (SCRSR) yields promising SR performance while maintaining shorter running time and fewer parameters.

Neural model of projecting flexural strength of cement concrete intended for airfield pavements

This work present to the mathematical model in the form of Artificial Neural Network (ANN), intended for projecting concrete flexural strength. Input data was classified according to the type of component material and its content in concrete mix (cement contents, coarse aggregate, fine aggregate, water and admixtures). In order to determine mathematical model, a multilayer, one-way perceptron network was used, recursion network with sigmoidal neurons. The model assumes that neurons are gathered in some layers (one input layer, hidden layers and one output layer). The conducted cross-section of the impact of variable parameters values (learning constant α and momentum values η) on the accuracy of representation of flexural strength was analyses. Assessment criterion was assumed as consideration the lowest mistake level and 100% compliance. According to the obtained results ANN was assumed to be the best representing network for constant value of momentum 0.3, learning constant of 0.04 and 9 neurons in a hidden layer. Very good coincidence of component models with experiment results was achieved. At testing stage, the coincidence was achieved at the level of 99.25%, in case of the assumed network structure. During model verification by means of experimental results, the average coincidence was 99.68%.

Single image super-resolution with multi-level feature fusion recursive network

Recursive learning can widen the receptive field of deep convolutional neural network while do not increase model parameters with parameters sharing. And dense skip connections can promote deep feature representation ability by reusing deep features in different receptive fields. Both techniques are very beneficial to improve performance in image restoration tasks. In this paper, we propose a new end to end deep network for single image super-resolution (SISR) by using both recursive residual feature extraction and multi-level features fusion, in which the multi-level deep features are firstly produced from the input low resolution (LR) image by recursive convolution units, and then fused to reconstruct high resolution (HR) image. The proposed scheme could achieve good super-resolution performance with relatively low complexity. Extensive experimental results on the benchmark tests verify the effectiveness of the proposed method.

Development and Application of the Metalloprotease Activity Multiplexed Bead-Based Immunoassay (MAMBI).

Metalloproteinases (MMPs) are zinc-dependent endopeptidases that cleave various proteins to regulate normal and diseased cellular functions, and as such, they play significant roles in human tissue development, homeostasis, and the pathogenesis of many diseases, including cancers, endometriosis, arthritis, etc. Most MMPs are produced as zymogenic latent enzymes that must be cleaved to activate their catalytic regions, and localized endogenous protein inhibitors further regulate activity. Accordingly, they operate within recursive networks to degrade extracellular matrix proteins and regulate cell signaling by cleaving growth factors and receptors at the cell surface and in the local pericellular environment. Thus, high-resolution information about the concentrations of specific active MMPs, revealing their intricate regulatory networks, may improve disease diagnosis and treatment. Here, we introduce a new and readily mastered method for measuring MMP activities in a multiplex fashion. We integrate aspects of activity-based enzyme labeling with commercial high-throughput, multiplexed protein quantification to yield the metalloproteinase activity multiplexed bead-based immunoassay (MAMBI). Assays of recombinant active MMP-1, -2, -3, -7, -8, -9, -12, and -13 establish the sensitivity and selectivity of MAMBI detection. Levels of active native MMPs are similarly characterized in conditioned cell culture medium, menstrual effluent, and uterine tissue. In a single MAMBI (5 μL), we achieve sensitivities equal to those from leading single-plex MMP activity detection strategies (e.g., 10-15 M for MMP-1). We also demonstrate high-throughput inhibitor screening via the MAMBI approach in complex, patient-derived samples.