Diffusion Tensor Magnetic Resonance Imaging (DTI) is a non-invasive technique for studying brain structure in vivo by measuring the diffusion properties of water molecules. Unlike conventional medical imaging that captures scalar intensity data, DTI data is typically stored as a 4D volume, where each voxel in 3D space is a 3×3 Cartesian tensor. DTI characterizes tensor-based diffusion profiles and captures information about the orientation of fiber bundles. During the alignment process, voxels need to be spatially transformed while maintaining the correspondence of tensor orientations, which leads to complex computations. Traditional DTI registration methods often suffer from slow iteration speed and low accuracy, posing challenges for clinical applications. In this paper, a novel DTI Registration method Based on Parallel Dual-channel Voxel Morph (DTI-RBPDV) is proposed. The core of the method is a two-branch convolutional neural network architecture. With a view to enhancing the alignment performance, it processes two input patterns simultaneously: (1) fractional anisotropy (FA) images and (2) principal eigenvectors from to-be-aligned and fixed DTI volumes to enhance the accuracy of deformation field prediction. In the network decoder layer, integration of attention mechanisms has also been implemented. These channel space attention modules dynamically highlight salient anatomical features and orientation consistency, improving the model's sensitivity to key structural alignments. Experimental results show that DTI-RBPDV effectively addresses the limitations of slow iterative computation and the challenges of applying deep learning to high-dimensional DTI data by significantly improving the registration accuracy and computational speed.

Due to the inherent inductive bias of operations, convolutional neural networks (CNN) cannot model global information of remote sensing (RS) images. In contrast, Transformer-based methods can establish long-range dependencies of images through self-attention (SA) mechanism, but it faces the challenges of computational complexity and memory requirements, but also ignores the exploration on the feature redundancy removal of RS images. To address these two issues, we propose a network based on dynamic sparse encoding and cross-temporal collaborative attention (DSECTCA-Net) for RS image change detection (CD). First, we implement dynamic sparse encoding (DSE) by designing hierarchical sparse Transformer module (HSTM), which decreases the correlation calculation of the SA mechanism and effectively reduces the computational complexity and parameter amount of Transformer. Secondly, we propose cross-temporal collaborative attention (CTCA) to model RS images in time series and fully explore the interactivity between dual-temporal RS images, so as to better extract the global understanding of visual scenes. Extensive experiments on two large-scale public RS datasets show that the proposed method not only provides higher detection accuracy, but also achieves lower computational complexity and required storage space than most popular CD networks.

Contrastive learning plays an important role in the semi-supervised medical image segmentation. However, existing contrastive learning methods struggle to capture the correlation of global and local features and improve feature discrimination for complex medical scenes, resulting in poor segmentation performance in challenging polyp segmentation. To overcome these limitations, we propose a semi-supervised polyp segmentation method using Hierarchical Granularity Contrastive Learning (HGCL). HGCL has two advantages. First, we design a hierarchical spatial contrastive learning module to divide the feature maps into large and small regions and perform different region-level contrastive learning, which can effectively capture the correlation of global and local information and improve the intra-class cohesion and inter-class separation. Second, we design a fine-granularity contrastive learning module, which can perform finer pixel-level contrastive learning to capture finer subtle local features and improve the generalization capacity of HGCL for complex medical scenes. Extensive experiments on three publicly available polyp datasets demonstrate that HGCL can achieve the better segmentation performance than existing popular semi-supervised methods. The code is available at https://github.com/Milk-White/HGCL.

Underwater images suffer from light absorption and scattering, impairs their visibility and applications. Existing underwater image restoration (UIR) methods based on generative models struggle are difficult to adapt to the complex and dynamic underwater environments characterized by illumination interference, low-light conditions, and non-uniform turbidity. To address these issues, we propose Water-CDM, a novel Adaptive Double-Branch Fusion Conditional Diffusion Model for underwater image restoration. Specifically, an adaptive double-branch fusion conditional diffusion model is presented utilizing a U-shaped full-attention network and Guided Multi-Scale Retinex with Brightness Correction (GMSRBC) to restore the challenging regions within underwater images. More precisely, to correct color casts and enhance the sharpness of underwater images, a U-shaped full-attention network incorporating Attention Blocks is designed for noise estimation during the reverse process of the conditional diffusion model. Concurrently, to mitigate overexposure during the enhancement of low-light underwater images under illumination interference, the GMSRBC method, featuring an Adaptive Brightness Correction Module, is proposed to efficiently adjust the brightness of underwater images. Experimental results demonstrate that the proposed Water-CDM significantly improves the quality of underwater images in challenging scenarios. Encouragingly, our proposed Water-CDM yields superior restoration outcomes compared to current state-of-the-art methods on three challenging publicly available datasets. Our codes will be released at: https://github.com/HKandWJJ/Water-CDM.

Intelligent anomaly detection (AD) methods have achieved much successes in machinery condition monitoring. However, the underlying independent and identically distributed assumption restricts their application scopes to steady operating conditions. False and missing alarms would occur when machines operate under time-varying circumstances. In this work, a more challenging time-varying setting is studied, where the working conditions are continuously changing, such that few or no samples are available for model training at one single condition. To tackle this issue, we propose a unified flowing normality learning (UFNL) framework, which aims to capture the flowing normal conditional distribution of time-varying samples and assigns dynamic decision boundary for AD. Specifically, a manifold-based probability density estimation is utilized to guide the adversarial learning process of generative adversarial networks, where adjacent samples are aggregated to approximate the conditional distribution by a conditional generator. Then, a latent normality inversion is proposed to extract the manifold structure from the pretrained generator and to map it into the latent space via a conditional encoder. The reconstruction errors from the encoder and generator can reveal the deviation of signals to the flowing normality. Finally, a condition-aware adaptive threshold selection strategy is proposed, where different thresholds are adaptively assigned for different conditions. Experiments are carried out under two typical continuous time-varying scenarios. The results demonstrate that the proposed framework can realize accurate fault detection at any operating condition within continuously changing environments.

The development in machine learning (ML) technology has brought new horizons for the prediction of subglacial lakes (SLs) using radio-echo sounding (RES) data, offering fresh perspectives toward the automated identification of SLs. Nonetheless, the inherent data imbalance across various classes within the dataset presents significant analytical challenges. To address this limitation, the artificial bee colony (ABC) optimization algorithm is introduced to automatically predict SLs in Gamburtsev Province in East Antarctica, using an optimized stacking ensemble learning approach. The proposed method predicts SLs by using five representative features selected through importance and correlation analyses of eight features derived from RES data. The experimental outcomes demonstrate the superiority of this method in overcoming the significant imbalance of RES data, successfully identifying known lakes in the validation dataset. Furthermore, this study summarizes an inventory of SLs across the Gamburtsev subglacial mountains in East Antarctica, and a total of 55 new candidate SLs with lengths ranging from 108 to 38130 m have been predicted using our novel method. The source code is publicly available athttps://github.com/vivian-ma97/ABC-Stacking-for-Subglacial-Lakes

Visual foundation models (VFMs) have been widely applied in the field of remote sensing (RS). However, they still face two main challenges when applied to precise remote sensing change detection (RSCD) tasks in complex scenes. Firstly, the nonnegligible domain shift between natural scene and RS scene limits the direct application of VFMs to the RSCD task. Second, most of existing RSCD methods may suffer from the boundary displacement problem due to the inadequate exploration of temporal differences for bi-temporal features. To address the above issues, this study proposes a SAM2-based domain adaptive and spatial difference aggregation network (DA2-Net) for RSCD. The proposed DA2-Net has two main advantages. First, a hierarchical low-rank adaptation (LoRA) strategy is presented by introducing low-rank matrices at key positions of SAM2, which can inject inductive biases from the RS domain into the network and alleviate the domain shift problem. Second, a difference adaptive enhancement module (DAEM) is designed to explore temporal differences for hierarchical bi-temporal features. The DAEM provides respective attention weights for different information through a dual branch of global difference awareness and local detail optimization. Experimental results on SYSU-CD, WHU-CD, and LEVIR-CD datasets demonstrate the superiority of DA2-Net. Code is available at https://github.com/xuptheqi-hash/ DA2Net.

Deep learning has attracted much attention in bearing fault diagnosis because of its high precision and end-to-end modules. However, in real industrial scenarios, some complex mechanical structures and working environments hinder data collection and fault reproduction, which makes bearing fault diagnosis with few samples a practical but challenging issue. As a data-driven approach, the standard deep learning method cannot extract features from a few samples due to overfitting. Neuroscience research has shown that the learning mechanism of the biological brain is more adaptable to learning tasks with few samples. Motivated by this, we propose a brain-inspired meta-learning (BIML) strategy for diagnosing few-shot bearing faults. Specifically, we design a brain-like learning algorithm for spiking neural networks (SNNs) based on the biological nervous system's learning mechanism and introduce a meta-learning strategy to apply it to the fault diagnosis task of bearing with few samples. Experimental results show that BIML is better than existing few-shot bearing fault diagnosis methods. Subsequently, we conduct a theoretical analysis of the effectiveness of BIML strategies and verify our analysis through experiments.

The following three factors restrict the application of existing low-light image enhancement methods: corruptions induced by the light-up process, color distortion, and a restricted generalization capacity due to limited paired training data. To address these limitations, we first combine HSV theory and Transformer, proposing a robust unsupervised low-light image enhancement framework, named HSVFormer. Secondly, we introduce brightness disturbance and design an unsupervised value enhancement network, which estimates brightness information and restores degraded brightness information to obtain enhanced reflectance. Finally, we utilize the V-subspace and devise a value-guided multi-head channel self-attention to capture brightness representations of regions with different brightness conditions and guide the modeling of non-local interactions. Experiment results on publicly available datasets demonstrate that HSVFormer can achieve superior performance compared with state-of-the-art approaches. The code is available at https://github.com/m0fig/HSVFormer.

Popular Transformer networks have been successfully applied to remote sensing (RS) image change detection (CD) identifications and achieved better results than most convolutional neural networks (CNNs), but they still suffer from two main problems. First, the computational complexity of the Transformer grows quadratically with the increase of image spatial resolution, which is unfavorable to RS images. Second, these popular Transformer networks tend to ignore the importance of fine-grained features, which results in poor edge integrity and internal tightness for largely changed objects and leads to the loss of small changed objects. To address the above issues, this letter proposes a lightweight structure-aware Transformer (LSAT) network for RS image CD. The proposed LSAT has two advantages. First, a cross-dimension interactive self-attention (CISA) module with linear complexity is designed to replace the vanilla self-attention (SA) in the visual Transformer, which effectively reduces the computational complexity while improving the feature representation ability of the proposed LSAT. Second, a structure-aware enhancement module (SAEM) is designed to enhance difference features and edge detail information, which can achieve double enhancement by difference refinement and detail aggregation to obtain fine-grained features of bi-temporal RS images. Experimental results show that the proposed LSAT achieves significant improvement in detection accuracy and offers a better tradeoff between accuracy and computational costs than most state-of-the-art (SOTA) CD methods for RS images.

The innovations in technologies have led to the emergence of digital finance such as online payment, online insurance, online lending, and supply chain finance. Digital finance has greatly facilitated people’s lives, accelerated the circulation of capital in various fields, and enhanced the vitality of financial markets. However, it exposes many increasing risks and hidden dangers such as stock volatility, trading fraud, credit card fraud, and privacy leakage[1],[2],[3],[4],[5],[6],[7]. How to effectively calculate, control, manage, and utilize financial big data and make full use of artificial intelligence technology to ensure financial security is an important research question. Solving it faces many challenges. These challenges not only include the complexity of data and computation but also the effectiveness of intelligent optimization algorithms and ways to deal with human behaviors and social environments[8],[9].

Remote sensing building change detection (RSBCD) detects changes in the spatial distribution of buildings which is of great significance for urban planning and construction. Existing deep learning-based RSBCD methods usually suffer from low object completeness and erroneous detection problem, mainly due to insufficient utilization of difference information between bi-temporal images. To address the above issues, this article proposed a new Siamese ensemble difference attention network (SEDANet) for RSBCD tasks in very-high resolution (VHR) images. Firstly, the key module ensemble difference attention module (EDAM) is designed to effectively extract difference representation between the bi-temporal features and filter out irrelevant changes. EDAM calculates difference map of bi-temporal features and transforms the extracted change information into trainable difference attention weights. The output weights from EDAM works as a guidance for both spatial and channel visual attention process, which enables the network to focus on foreground building changes and further resolve erroneous attention problems in existing RSBCD methods. The Siamese structure is adopted to better represent bi-temporal features, and convolutional blocks are replaced with residual convolution blocks (RCBs) to speed up network fitting and prevent gradient explosion or descent. We conduct comprehensive experiments on three benchmark datasets. Both visual and quantitative results show that our proposed SEDANet is superior to other eight state-of-the-art networks. Especially on GZ-CD dataset, SEDANet outperforms other comparison methods by 3%-8%. In addition, the effectiveness of EDAM module is also discussed through a series of ablation studies.

There are relatively few studies on the multi-coil reconstruction task of existing Magnetic Resonance Imaging (MRI) methods, as there are problems with insufficient reconstruction details, high memory occupation during training, etc. Therefore, a new Dual-domain Parallel Fusion Reconstruction Network (DPFNet) is proposed in this paper. The whole network consists of coil sensitivity graph estimation module, dual domain feature extraction module, dual domain dynamic error correction module, and dual domain dynamic fusion module. A U-Net has been used as the backbone network. The network reconstructs under-sampled MRI images and K-space data simultaneously in two branches of the image domain and K-space domain, and the fusion module realizes the reconstruction information interaction between the two branches. In addition, a new dual domain consistency loss is also proposed, which reduces the error between the same MRI slice image and K-space data with dual domain output, and achieves high quality reconstruction. In this paper, a series of comparative experiments and ablation experiments are conducted in the open Calgary-Campinas-359 brain MRI data set. The results of the experiments show that the proposed DPFNet achieves the most advanced level at present and is superior to other traditional algorithms and reconstruction methods based on deep learning. In particular, the reconstruction results from Cartesian sampling are very good.

The popular networks for change detection (CD) in very-high-resolution (VHR) remote sensing (RS) images usually suffer from two problems. First, it is difficult for these networks to model simultaneously the local and global features of changed targets, which leads to the limited feature representation ability of popular CD networks. Second, these networks often have a large number of parameters and high computational costs due to complex network architecture. To address the above issues, we propose a local-global siamese network (LGS-Net) for CD in VHR RS images. First, we design an encoder with a parallel dual-branch structure consisting of convolutional neural networks (CNNs) and Transformer to extract rich features from bi-temporal images. Furthermore, we design a local-global feature enhancement (LGFE) module to help our encoder improve its feature representation ability. Second, we design a compact and efficient convolution module called inter-scale separable convolution (ISSConv). This module first divides feature maps into multiple groups, and then performs depthwise separable convolution in each group using atrous convolution with different dilation rates, which can not only capture changed targets across scales but also effectively reduce the number of model parameters. Experiments demonstrate that the proposed LGS-Net is superior to the state-of-the-art CD networks in terms of parameters, computational costs, and detection accuracy.

Polyp segmentation is of great importance for the diagnosis and treatment of colorectal cancer. However, it is difficult to segment polyps accurately due to a large number of tiny polyps and the low contrast between polyps and the surrounding mucosa. To address this issue, we design an Adaptive Tiny-object Enhanced Network (ATENet) for tiny polyp segmentation. The proposed ATENet has two advantages: First, we design an adaptive tiny-object encoder containing three parallel branches, which can effectively extract the shape and position features of tiny polyps and thus improve the segmentation accuracy of tiny polyps. Second, we design a simple enhanced feature decoder, which can not only suppress the background noise of feature maps, but also supplement the detail information to improve further the polyp segmentation accuracy. Extensive experiments on three benchmark datasets demonstrate that the proposed ATENet can achieve the state-of-the-art performance while maintaining low computational complexity.

The hybrid architecture of convolutional neural networks (CNNs) and Transformer are very popular for medical image segmentation. However, it suffers from two challenges. First, although a CNNs branch can capture the local image features using vanilla convolution, it cannot achieve adaptive feature learning. Second, although a Transformer branch can capture the global features, it ignores the channel and cross-dimensional self-attention, resulting in a low segmentation accuracy on complex-content images. To address these challenges, we propose a novel hybrid architecture of convolutional neural networks hand in hand with vision Transformers (CiT-Net) for medical image segmentation. Our network has two advantages. First, we design a dynamic deformable convolution and apply it to the CNNs branch, which overcomes the weak feature extraction ability due to fixed-size convolution kernels and the stiff design of sharing kernel parameters among different inputs. Second, we design a shifted-window adaptive complementary attention module and a compact convolutional projection. We apply them to the Transformer branch to learn the cross-dimensional long-term dependency for medical images. Experimental results show that our CiT-Net provides better medical image segmentation results than popular SOTA methods. Besides, our CiT-Net requires lower parameters and less computational costs and does not rely on pre-training. The code is publicly available at https://github.com/SR0920/CiT-Net.

Phages are the functional viruses that infect bacteria and they play important roles in microbial communities and ecosystems. Phage research has attracted great attention due to the wide applications of phage therapy in treating bacterial infection in recent years. Metagenomics sequencing technique can sequence microbial communities directly from an environmental sample. Identifying phage sequences from metagenomic data is a vital step in the downstream of phage analysis. However, the existing methods for phage identification suffer from some limitations in the utilization of the phage feature for prediction, and therefore their prediction performance still need to be improved further. In this article, we propose a novel deep neural network (called MetaPhaPred) for identifying phages from metagenomic data. In MetaPhaPred, we first use a word embedding technique to encode the metagenomic sequences into word vectors, extracting the latent feature vectors of DNA words. Then, we design a deep neural network with a convolutional neural network (CNN) to capture the feature maps in sequences, and with a bi-directional long short-term memory network (Bi-LSTM) to capture the long-term dependencies between features from both forward and backward directions. The feature map consists of a set of feature patterns, each of which is the weighted feature extracted by a convolution filter with convolution kernels in the CNN slide along the input feature vectors. Next, an attention mechanism is used to enhance contributions of important features. Experimental results on both simulated and real metagenomic data with different lengths demonstrate the superiority of the proposed MetaPhaPred over the state-of-the-art methods in identifying phage sequences.

Deep convolutional neural networks have achieved much success in remote sensing image change detection (CD) but still suffer from two main problems. First, existing multi-scale feature fusion methods often employ redundant feature extraction and fusion strategies, which often leads to high computational costs and memory usage. Second, the regular attention mechanism in CD is difficult to model spatial-spectral features and generate 3D attention weights at the same time, ignoring the cooperation between spatial features and spectral features. To address the above issues, an efficient ultra-lightweight spatial-spectral feature cooperation network (USSFC-Net) is proposed for CD in this paper. The proposed USSFC-Net has two main advantages. First, a multi-scale decoupled convolution (MSDConv) is designed, which is clearly different from the popular atrous spatial pyramid pooling (ASPP) module and its variants since it can flexibly capture the multi-scale features of changed objects by using cyclic multi-scale convolution. Meanwhile, the design of MSDConv can greatly reduce the number of parameters and computational redundancy. Second, an efficient spatial-spectral feature cooperation strategy (SSFC) is introduced to obtain richer features. The SSFC differs from existing 2D attention mechanisms since it learns 3D spatial-spectral attention weights without adding any parameters. The experiments on three datasets for remote sensing image CD demonstrate that the proposed USSFC-Net achieves better CD accuracy than most convolutional neural networks-based methods and requires lower computational costs and fewer parameters, even it is superior to some Transformer-based methods. The code is available at https://github.com/SUST-reynole/USSFC-Net.

Although deep learning-based change detection (CD) methods achieve great success in remote sensing images, they still suffer from two main challenges. First, popular Convolutional Neural Networks (CNNs) are weak in extracting discriminated features focusing on changed regions, since most methods ignore the multi-frequency components of bi-temporal images. Second, although existing CD methods employ the Transformer structure to capture long-range dependency for global feature representation, it is difficult for them to simultaneously take into account the long-range dependency of changed objects at various scales. To address the above issues, we propose a triple change detection network (TCD-Net) via joint multi-frequency and full-scale Swin-Transformer. The proposed TCD-Net has two main advantages. First, we propose a multi-frequency channel attention (MFCA) module to boost the ability of modeling the channel correlation, which can compensate for the problem of insufficient feature representation caused by only performing global average pooling (GAP). Furthermore, a joint multi-frequency difference feature enhancement (JM-DFE) guiding block is proposed to improve the boundary quality and the position awareness of truly changed objects, which can effectively extract channel features of multi-frequency information and thus improve the discriminative ability of features. Second, unlike Siamese-based structures, we propose a full-scale Swin-Transformer (FST) module as the third branch to model and aggregate the long-range dependency of multi-scale changed objects, which can alleviate the missed detections of small objects and achieve more compact changed regions effectively. Experiments on three public CD datasets exhibit that the proposed TCD-Net achieves better CD accuracy with smaller model complexity than state-of-the-art methods. The code is publicly available at https://github.com/RSCD-mz/TCD-Net.

Convolutional neural networks (CNNs) have achieved significant success in medical image segmentation. However, they also suffer from the requirement of a large number of parameters, leading to a difficulty of deploying CNNs to low-source hardwares, e.g., embedded systems and mobile devices. Although some compacted or small memory-hungry models have been reported, most of them may cause degradation in segmentation accuracy. To address this issue, we propose a shape-guided ultralight network (SGU-Net) with extremely low computational costs. The proposed SGU-Net includes two main contributions: it first presents an ultralight convolution that is able to implement double separable convolutions simultaneously, i.e., asymmetric convolution and depthwise separable convolution. The proposed ultralight convolution not only effectively reduces the number of parameters but also enhances the robustness of SGU-Net. Secondly, our SGU-Net employs an additional adversarial shape-constraint to let the network learn shape representation of targets, which can significantly improve the segmentation accuracy for abdomen medical images using self-supervision. The SGU-Net is extensively tested on four public benchmark datasets, LiTS, CHAOS, NIH-TCIA and 3Dircbdb. Experimental results show that SGU-Net achieves higher segmentation accuracy using lower memory costs, and outperforms state-of-the-art networks. Moreover, we apply our ultralight convolution into a 3D volume segmentation network, which obtains a comparable performance with fewer parameters and memory usage.

Popular semi-supervised medical image segmentation networks often suffer from error supervision from unlabeled data since they usually use consistency learning under different data perturbations to regularize model training. These networks ignore the relationship between labeled and unlabeled data, and only compute single pixel-level consistency leading to uncertain prediction results. Besides, these networks often require a large number of parameters since their backbone networks are designed depending on supervised image segmentation tasks. Moreover, these networks often face a high over-fitting risk since a small number of training samples are popular for semi-supervised image segmentation. To address the above problems, in this paper, we propose a novel adversarial self-ensembling network using dynamic convolution (ASE-Net) for semi-supervised medical image segmentation. First, we use an adversarial consistency training strategy (ACTS) that employs two discriminators based on consistency learning to obtain prior relationships between labeled and unlabeled data. The ACTS can simultaneously compute pixel-level and image-level consistency of unlabeled data under different data perturbations to improve the prediction quality of labels. Second, we design a dynamic convolution-based bidirectional attention component (DyBAC) that can be embedded in any segmentation network, aiming at adaptively adjusting the weights of ASE-Net based on the structural information of input samples. This component effectively improves the feature representation ability of ASE-Net and reduces the overfitting risk of the network. The proposed ASE-Net has been extensively tested on three publicly available datasets, and experiments indicate that ASE-Net is superior to state-of-the-art networks, and reduces computational costs and memory overhead. The code is available at: https://github.com/SUST-reynole/ASE-Nethttps://github.com/SUST-reynole/ASE-Net.

The popular convolutional neural networks (CNNs) have been successfully used in very high-resolution remote sensing image semantic segmentation. However, these networks often suffer from performance limitations. First, although deeper networks usually provide better feature representation, they may cause parameter redundancy and the inefficient use of prior knowledge. Secondly, attention-based networks often only focus on weighting different features of a single sample but ignore the correlation of all samples in training set, thus leading to the loss of global information. To address above issues, we propose two simple yet effective global evolution strategies. The first is knowledge enhancement. This strategy can reactivate invalid convolutional kernels through convergence of different models and make full use of prior knowledge from the network to improve its feature representation. The second is a dict-attention module that greatly enhances the generalization of networks by learning and inferring the global relationship among different samples through the dictionary unit. As a result, a novel global evolution network (GENet) is designed based on knowledge enhancement and dict-attention for remote sensing image semantic segmentation. Experiments demonstrate that the proposed GENet is not only superior to popular networks in segmentation accuracy.

Abstract Deep learning has been widely used for medical image segmentation and a large number of papers has been presented recording the success of deep learning in the field. A comprehensive thematic survey on medical image segmentation using deep learning techniques is presented. This paper makes two original contributions. Firstly, compared to traditional surveys that directly divide literatures of deep learning on medical image segmentation into many groups and introduce literatures in detail for each group, we classify currently popular literatures according to a multi‐level structure from coarse to fine. Secondly, this paper focuses on supervised and weakly supervised learning approaches, without including unsupervised approaches since they have been introduced in many old surveys and they are not popular currently. For supervised learning approaches, we analyse literatures in three aspects: the selection of backbone networks, the design of network blocks, and the improvement of loss functions. For weakly supervised learning approaches, we investigate literature according to data augmentation, transfer learning, and interactive segmentation, separately. Compared to existing surveys, this survey classifies the literatures very differently from before and is more convenient for readers to understand the relevant rationale and will guide them to think of appropriate improvements in medical image segmentation based on deep learning approaches.

Fuzzy c-means (FCM) algorithms with spatial information have been widely applied in the field of image segmentation. However, most of them suffer from two challenges. One is that the introduction of fixed or adaptive single neighboring information with narrow receptive field limits contextual constraints leading to clutter segmentations. The other is that the incorporation of superpixels with wide receptive field enlarges spatial coherency leading to block effects. To address these challenges, we propose fuzzy STUDENT’S t-distribution model based on richer spatial combination (FRSC) for image segmentation. In this article, we make two significant contributions. The first is that both the narrow and wide receptive fields are integrated into the objective function of FRSC, which is convenient to mine image features and distinguish local difference. The second is that the rich spatial combination under STUDENT’S t-distribution ensures that spatial information is introduced into the updated parameters of FRSC, which is helpful in finding a balance between the noise-immunity and detail-preservation. Experimental results on synthetic and publicly available images further demonstrate that the proposed FRSC addresses successfully the limitations of FCM algorithms with spatial information, and provides better segmentation results than state-of-the-art clustering algorithms.

Auscultation is an important tool for diagnosing respiratory-related diseases. Unfortunately, the quality of auscultation is limited by the professional level of the doctor and the environment of the auscultation. Some studies have focused on automated auscultation techniques. However, existing approaches suffer from two challenges: 1) the models cannot learn from data distributed among multiple hospitals and 2) the predictions of the models are difficult to interpret for physicians. To address this issue, this article proposes a novel explainable respiratory sound analysis framework with fuzzy decision tree regularization. This framework develops an ensemble knowledge distillation technique to learn distributed data and achieves good performance in terms of model efficiency and accuracy. Fuzzy decision trees are used to explain the predictions of the model and produce decision rules that can be well accepted by physicians. The effectiveness of this framework is thoroughly validated on the Respiratory Sound database and compared with other existing approaches.

The popular Siamese convolutional neural networks (CNNs) for remote sensing (RS) image change detection (CD) often suffer from two problems. First, they either ignore the original information of bitemporal images or insufficiently utilize the difference information between bitemporal images, which leads to the low tightness of the changed objects. Second, Siamese CNNs always employ dual-branch encoders for CD, which increases computational cost. To address the above issues, this article proposes a network based on difference enhancement and spatial–spectral nonlocal (DESSN) for CD in very-high-resolution (VHR) images. This article makes threefold contributions. First, we design a difference enhancement (DE) module that can effectively learn the difference representation between foreground and background to reduce the impact of irrelevant changes on the detection results. Second, we present a spatial–spectral nonlocal (SSN) module that is different from vanilla nonlocal because multiscale spatial global features are incorporated to model the large-scale variation of objects during CD. The module can be used to strengthen the edge integrity and internal tightness of changed objects. Third, the asymmetric double convolution with Ghost (ADCG) module is exploited instead of standard convolution. The ADCG can not only refine the edge information of the changed objects, since horizontal and vertical convolutional kernels have good contour preservation advantages, but also greatly reduce the computational complexity of the proposed model. The experiments on two public VHR CD datasets demonstrate that the proposed network can provide higher detection accuracy and requires smaller memory usage than state-of-the-art networks.

Stroke has become a leading cause of death and long-term disability in the world with no effective treatment. Deep learning-based approaches have the potential to outperform existing stroke risk prediction models, but they rely on large well-labeled data. Due to the strict privacy protection policy in health-care systems, stroke data is usually distributed among different hospitals in small pieces. In addition, the positive and negative instances of such data are extremely imbalanced. Transfer learning can solve small data issue by exploiting the knowledge of a correlated domain, especially when multiple source of data are available. In this work, we propose a novel Hybrid Deep Transfer Learning-based Stroke Risk Prediction (HDTL-SRP) scheme to exploit the knowledge structure from multiple correlated sources (i.e., external stroke data, chronic diseases data, such as hypertension and diabetes). The proposed framework has been extensively tested in synthetic and real-world scenarios, and it outperforms the state-of-the-art stroke risk prediction models. It also shows the potential of real-world deployment among multiple hospitals aided with 5 G/B5G infrastructures.

Many real systems are represented in form of multiplex networks composed of a set of nodes, multiple layers of links, and coupling node relationships across all layers. These systems are very vulnerable to damages during both attacks and recoveries due to potential node cascading failures (NCFs). Although some progress has recently been made in studying network robustness and resilience, the comprehensive impacts of coupling node relationships and community structures on NCFs remain unclear. Accordingly, in this article, we study the robustness and resilience of multiplex networks in the presence of NCFs caused by coupling node relationships and community structures. We first model the failure processes of multiplex networks during both attacks and recoveries as node-community cascading failures (called NCCFs), and then theoretically demonstrate the fragility of multiplex networks to random node damages under NCCFs. Subsequently, to improve network robustness and resilience, we adopt a node protection strategy and propose a cost-aware constrained optimization problem. Finally, we devise a degree-based simulated annealing algorithm for solving this optimization problem. Extensive experiments on both simulated and real multiplex networks show that NCCFs make networks more vulnerable to unpredictable damage than classical NCFs. The results also show the superiority of the proposed algorithm over the state-of-the-art algorithms in improving network robustness and resilience.

The popular deep convolutional networks used for image super-resolution (SR) reconstruction often increase the network depth and employ attention mechanism to improve image reconstruction effect. However, these networks suffer from two problems. The first is the deeper network easily causes higher computational cost and more GPU memory usage. The second is traditional attention mechanism often misses the spatial information of images leading the loss of image detail information. To address these issues, we propose a lightweight non-local network (LNLN) for image super resolution in this paper. The proposed network makes two contributions. First, we use non-local module instead of normal attention module to obtain larger receptive field and extract more comprehensive feature information, which is helpful for improving image SR reconstruction results. Secondly, we use the depthwise separable convolution (DSC) instead of the vanilla convolution to reconstruct the residual block, which greatly reduces the number of parameters and computational cost. The proposed LNLN and comparative networks are evaluated on five commonly public datasets, and experiments demonstrate that the proposed LNLN is superior to state-of-the-art networks in terms of reconstruction performance, the number of parameters and storage space.

Abstract Although deep learning has been widely used for dense crowd counting, it still faces two challenges. Firstly, the popular network models are sensitive to scale variance of human head, human occlusions, and complex background due to repeated utilization of vanilla convolution kernels. Secondly, the vanilla feature fusion often depends on summation or concatenation, which ignores the correlation of different features leading to information redundancy and low robustness to background noise. To address these issues, a multi‐scale feature pyramid network (MFP‐Net) for dense crowd counting is proposed in this paper. The proposed MFP‐Net makes two contributions. Firstly, the feature pyramid fusion module is designed that adopts rich convolutions with different depths and scales, not only to expand the receptive field, but also to improve the inference speed of models by using parallel group convolution. Secondly, a feature attention‐aware module is added in the feature fusion stage. The module can achieve local and global information fusion by capturing the importance of the spatial and channel domains to improve model robustness. The proposed MFP‐Net is evaluated on five publicly available datasets, and experiments show that the MFP‐Net not only provides better crowd counting results than comparative models, but also requires fewer parameters.

Discovering DNA-protein binding sites, also known as motif discovery, is the foundation for further analysis of transcription factors (TFs). Deep learning algorithms such as convolutional neural networks (CNN) have been introduced to motif discovery task and have achieved state-of-art performance. However, due to the limitations of CNN, motif discovery methods based on CNN do not take full advantage of large-scale sequencing data generated by high-throughput sequencing technology. Hence, in this paper we propose multi-scale capsule network architecture (MSC) integrating multi-scale CNN, a variant of CNN able to extract motif features of different lengths, and capsule network, a novel type of artificial neural network architecture aimed at improving CNN. The proposed method is tested on real ChIP-seq datasets and the experimental results show a considerable improvement compared with two well-tested deep learning-based sequence model, DeepBind and Deepsea.

The V-Net based 3D fully convolutional neural networks have been widely used in liver volumetric data segmentation. However, due to the large number of parameters of these networks, 3D FCNs suffer from high computational cost and GPU memory usage. To address these issues, we design a lightweight V-Net (LV-Net) for liver segmentation in this paper. The proposed network makes two contributions. The first is that we design an inverted residual bottleneck block (IRB block) and a 3D average pooling block and apply them to the proposed LV-Net. Compared with vanilla convolution, depth-wise convolution and point-wise convolution employed by the IRB block can not only reduce the number of parameters significantly, but also extract features sufficiently well by decoupling cross-channel corrections and spatial correlations. The second is that the LV-Net employs 3D deep supervision to improve the final loss function in training phase, which makes the proposed LV-Net acquire a more powerful discrimination capability between liver areas and non-liver areas. The proposed LV-Net is evaluated on public LiTS dataset, and experiments demonstrate that the proposed LV-Net is superior to popular 2D and 3D networks in terms of segmentation performance, parameter quantity and computational cost.

Clustering algorithms by minimizing an objective function share a clear drawback of having to set the number of clusters manually. Although density peak clustering is able to find the number of clusters, it suffers from memory overflow when it is used for image segmentation because a moderate-size image usually includes a large number of pixels leading to a huge similarity matrix. To address this issue, here we proposed an automatic fuzzy clustering framework (AFCF) for image segmentation. The proposed framework has threefold contributions. First, the idea of superpixel is used for the density peak (DP) algorithm, which efficiently reduces the size of the similarity matrix and thus improves the computational efficiency of the DP algorithm. Second, we employ a density balance algorithm to obtain a robust decision-graph that helps the DP algorithm achieve fully automatic clustering. Finally, a fuzzy c-means clustering based on prior entropy is used in the framework to improve image segmentation results. Because the spatial neighboring information of both the pixels and membership are considered, the final segmentation result is improved effectively. Experiments show that the proposed framework not only achieves automatic image segmentation, but also provides better segmentation results than state-of-the-art algorithms.

In this paper, we propose a novel approach based on a symmetric fully convolutional network within pyramid pooling (FCN-PP) for landslide mapping (LM). The proposed approach has three advantages. Firstly, this approach is automatic and insensitive to noise because multivariate morphological reconstruction (MMR) is used for image preprocessing. Secondly, it is able to take into account features from multiple convolutional layers and explore efficiently the context of images, which leads to a good tradeoff between wider receptive field and the use of context. Finally, the selected pyramid pooling module addresses the drawback of single-scale pooling employed by convolutional neural network (CNN), fully convolutional network (FCN), U-Net, etc. Experimental results show that the proposed FCN-PP is effective for LM, and it outperforms state-of-the-art approaches in terms of four metrics, Precision, Recall, F -score, and Accuracy.

Most of the approaches used for Landslide inventory mapping (LIM) rely on traditional feature extraction and unsupervised classification algorithms. However, it is difficult to use these approaches to detect landslide areas because of the complexity and spatial uncertainty of landslides. In this letter, we propose a novel approach based on a fully convolutional network within pyramid pooling (FCN-PP) for LIM. The proposed approach has three advantages. First, this approach is automatic and insensitive to noise because multivariate morphological reconstruction is used for image preprocessing. Second, it is able to take into account features from multiple convolutional layers and explore efficiently the context of images, which leads to a good tradeoff between wider receptive field and the use of context. Finally, the selected PP module addresses the drawback of global pooling employed by convolutional neural network, FCN, and U-Net, and, thus, provides better feature maps for landslide areas. Experimental results show that the proposed FCN-PP is effective for LIM, and it outperforms the state-of-the-art approaches in terms of five metrics, $Precision$ , $Recall$ , $Overall~Error$ , $F$ -$score$ , and $Accuracy$ .

Underlying a cancer phenotype is a specific gene regulatory network that represents the complex regulatory relationships between genes. It remains, however, a challenge to find cancer-related gene regulatory network because of insufficient sample sizes and complex regulatory mechanisms in which gene is influenced by not only other genes but also other biological factors. With the development of high-throughput technologies and the unprecedented wealth of multi-omics data it gives us a new opportunity to design machine learning method to investigate underlying gene regulatory network. In this paper, we propose an approach, which use Biweight Midcorrelation to measure the correlation between factors and make use of Nonconvex Penalty based sparse regression for Gene Regulatory Network inference (BMNPGRN). BMNCGRN incorporates multi-omics data (including DNA methylation and copy number variation) and their interactions in gene regulatory network model. The experimental results on synthetic datasets show that BMNPGRN outperforms popular and state-of-the-art methods (including DCGRN, ARACNE, and CLR) under false positive control. Furthermore, we applied BMNPGRN on breast cancer (BRCA) data from The Cancer Genome Atlas database and provided gene regulatory network.

A great number of improved fuzzy c-means (FCM) clustering algorithms have been widely used for grayscale and color image segmentation. However, most of them are time-consuming and unable to provide desired segmentation results for color images due to two reasons. The first one is that the incorporation of local spatial information often causes a high computational complexity due to the repeated distance computation between clustering centers and pixels within a local neighboring window. The other one is that a regular neighboring window usually breaks up the real local spatial structure of images and thus leads to a poor segmentation. In this work, we propose a superpixel-based fast FCM clustering algorithm that is significantly faster and more robust than stateof-the-art clustering algorithms for color image segmentation. To obtain better local spatial neighborhoods, we first define a multiscale morphological gradient reconstruction operation to obtain a superpixel image with accurate contour. In contrast to traditional neighboring window of fixed size and shape, the superpixel image provides better adaptive and irregular local spatial neighborhoods that are helpful for improving color image segmentation. Second, based on the obtained superpixel image, the original color image is simplified efficiently and its histogram is computed easily by counting the number of pixels in each region of the superpixel image. Finally, we implement FCM with histogram parameter on the superpixel image to obtain the final segmentation result. Experiments performed on synthetic images and real images demonstrate that the proposed algorithm provides better segmentation results and takes less time than state-of-the-art clustering algorithms for color image segmentation.

Morphological reconstruction (MR) is often employed by seeded image segmentation algorithms such as watershed transform and power watershed, as it is able to filter out seeds (regional minima) to reduce over-segmentation. However, the MR might mistakenly filter meaningful seeds that are required for generating accurate segmentation and it is also sensitive to the scale because a single-scale structuring element is employed. In this paper, a novel adaptive morphological reconstruction (AMR) operation is proposed that has three advantages. First, AMR can adaptively filter out useless seeds while preserving meaningful ones. Second, AMR is insensitive to the scale of structuring elements because multiscale structuring elements are employed. Finally, the AMR has two attractive properties: monotonic increasingness and convergence that help seeded segmentation algorithms to achieve a hierarchical segmentation. Experiments clearly demonstrate that the AMR is useful for improving performance of algorithms of seeded image segmentation and seed-based spectral segmentation. Compared to several state-of-the-art algorithms, the proposed algorithms provide better segmentation results requiring less computing time.

Summary In recent years, researchers have been trying to detect human emotions from recorded brain signals such as electroencephalogram (EEG) signals. However, due to the high levels of noise from the EEG recordings, a single feature alone cannot achieve good performance. A combination of distinct features is the key for automatic emotion detection. In this paper, we present a hybrid dimension feature reduction scheme using a total of 14 different features extracted from EEG recordings. The scheme combines these distinct features in the feature space using both supervised and unsupervised feature selection processes. Maximum Relevance Minimum Redundancy (mRMR) is applied to re‐order the combined features into max‐relevance with the labels and min‐redundancy of each feature. The generated features are further reduced with principal component analysis (PCA) for extracting the principal components. Experimental results show that the proposed work outperforms the state‐of‐art methods using the same settings in the publicly available DEAP data set.

Post translational modification plays a significiant role in the biological processing. The potential post translational modification is composed of the center sites and the adjacent amino acid residues which are fundamental protein sequence residues. It can be helpful to perform their biological functions and contribute to understanding the molecular mechanisms that are the foundations of protein design and drug design. The existing algorithms of predicting modified sites often have some shortcomings, such as lower stability and accuracy. In this paper, a combination of physical, chemical, statistical, and biological properties of a protein have been ulitized as the features, and a novel framework is proposed to predict a protein's post translational modification sites. The multi-layer neural network and support vector machine are invoked to predict the potential modified sites with the selected features that include the compositions of amino acid residues, the E-H description of protein segments, and several properties from the AAIndex database. Being aware of the possible redundant information, the feature selection is proposed in the propocessing step in this research. The experimental results show that the proposed method has the ability to improve the accuracy in this classification issue.

As fuzzy c-means clustering (FCM) algorithm is sensitive to noise, local spatial information is often introduced to an objective function to improve the robustness of the FCM algorithm for image segmentation. However, the introduction of local spatial information often leads to a high computational complexity, arising out of an iterative calculation of the distance between pixels within local spatial neighbors and clustering centers. To address this issue, an improved FCM algorithm based on morphological reconstruction and membership filtering (FRFCM) that is significantly faster and more robust than FCM is proposed in this paper. First, the local spatial information of images is incorporated into FRFCM by introducing morphological reconstruction operation to guarantee noise-immunity and image detail-preservation. Second, the modification of membership partition, based on the distance between pixels within local spatial neighbors and clustering centers, is replaced by local membership filtering that depends only on the spatial neighbors of membership partition. Compared with state-of-the-art algorithms, the proposed FRFCM algorithm is simpler and significantly faster, since it is unnecessary to compute the distance between pixels within local spatial neighbors and clustering centers. In addition, it is efficient for noisy image segmentation because membership filtering are able to improve membership partition matrix efficiently. Experiments performed on synthetic and real-world images demonstrate that the proposed algorithm not only achieves better results, but also requires less time than the state-of-the-art algorithms for image segmentation.

Owing to the importance of rolling element bearings in rotating machines, condition monitoring of rolling element bearings has been studied extensively over the past decades. However, most of the existing techniques require large storage and time for signal processing. This paper presents a new strategy based on compressive sensing for bearing faults classification that uses fewer measurements. Under this strategy, to match the compressed sensing mechanism, the compressed vibration signals are first obtained by resampling the acquired bearing vibration signals in the time domain with a random Gaussian matrix using different compressed sensing sampling rates. Then three approaches have been chosen to process these compressed data for the purpose of bearing fault classification these includes using the data directly as the input of classifier, and extract features from the data using linear feature extraction methods, namely, unsupervised Principal Component Analysis (PCA) and supervised Linear Discriminant Analysis (LDA). Classification performance using Logistic Regression Classifier (LRC) achieved high classification accuracy with significantly reduced bandwidth consumption compared with the existing techniques.

Rolling element bearings have a pivotal role in rotating machine and their failures are the leading cause of more substantial failures in the machine. In response to their importance, there is a growing body of research looking at condition monitoring of rolling element bearings to avoid machine breakdowns. In this study, by taking advantages of Compressive Sampling (CS), Laplacian Score (LS) and Multi-class Support Vector Machine (MSVM), an intelligent method for rolling bearing fault classification is proposed The CS is used to obtain compressed samples of the raw vibration signals, and the LS is used to rank the features of the obtained compressed samples with respect to their importance and correlations with the core fault characteristics. Then, based on LS ranking, we selected a small amount of the most significant compressed samples to produce the features vector. Finally, classification performance using MSVM shows high classification accuracy with a significantly reduced feature set.

In the past decades, neuroimaging of humans has gained a position of status within neuroscience, and data-driven approaches and functional connectivity analyses of functional magnetic resonance imaging (fMRI) data are increasingly favored to depict the complex architecture of human brains. However, the reliability of these findings is jeopardized by too many analysis methods and sometimes too few samples used, which leads to discord among researchers. We propose a tunable consensus clustering paradigm that aims at overcoming the clustering methods selection problem as well as reliability issues in neuroimaging by means of first applying several analysis methods (three in this study) on multiple datasets and then integrating the clustering results. To validate the method, we applied it to a complex fMRI experiment involving affective processing of hundreds of music clips. We found that brain structures related to visual, reward, and auditory processing have intrinsic spatial patterns of coherent neuroactivity during affective processing. The comparisons between the results obtained from our method and those from each individual clustering algorithm demonstrate that our paradigm has notable advantages over traditional single clustering algorithms in being able to evidence robust connectivity patterns even with complex neuroimaging data involving a variety of stimuli and affective evaluations of them. The consensus clustering method is implemented in the R package "UNCLES" available on http://cran.r-project.org/web/packages/UNCLES/index.html .

Automatic fault detection and classification for roller element bearings is an important issue for rotating machine condition monitoring. In this paper, we classify roller element bearings fault classes under two and three hidden layers' deep neural network framework based on sparse Autoencoder. This allows us to learn and extract features for the bearing vibration samples in an unsupervised manner using the encoder part of the Autoencoder. Then we form the deep neural network by stacking the encoders in each stage of the hidden layers together with the softmax layer. Classification performance using the full deep network and backpropagation compared, and effects of different deep neural network parameters on the classification accuracy are studied here.

In this paper, we propose a new consensus clustering algorithm, which is based on an existing clustering paradigm, called enhanced splitting merging awareness tactics (E-SMART). The problem of determining the number of clusters, which affects many state-of-theart consensus clustering algorithms, is addressed by the proposed CoCE-SMART algorithm. The idea behind CoCE-SMART is that SMART is used repeatedly to one dataset, resulting in different clustering results, which might have different numbers of clusters. These SMART clustering results can be combined by clustering the centroids of all clusters as the estimate of real number of clusters can be determined from the SMART clustering results. Three benchmark datasets are utilised to assess the proposed algorithm. The experimental results strongly indicate that the proposed CoCE-SMART algorithm outperforms other state-of-the-art consensus clustering algorithms.

In this paper, we propose a scalable clustering paradigm to address the problems of excessive computational load and limited clustering performance in large-scale data. The proposed method employs the enhanced splitting merging awareness tactics (E-SMART) algorithm. The large-scale dataset is divided into many sub-datasets sampled randomly from original data. These sub-datasets are clustered using E-SMART with the number of clusters K detected automatically and the resulting partitions are combined and re-clustered. We evaluate our method using synthetic fMRI datasets with different noise levels and one real fMRI dataset. Results show that the accuracy and execution time outperforms the traditional clustering algorithms in large-scale datasets.

Non-data aided channel estimation is discussed in this paper to enable blind modulation classification in multiple-input multiple-output fading channels. The channel parameters are jointly estimated via expectation maximization under each modulation hypothesis. Instead of pilot symbols, the initialization of the channel matrix is achieved through a combination of fuzzy c-means clustering and maximum likelihood mapping. The estimated channel matrix and noise power enable the blind classification of modulations using a maximum likelihood classifier. Digital modulations are tested in simulation to validate the proposed classifier. The classifier is able to achieve excellent performance when SNR level is above 5 dB.

BACKGROUND: Collective analysis of the increasingly emerging gene expression datasets are required. The recently proposed binarisation of consensus partition matrices (Bi-CoPaM) method can combine clustering results from multiple datasets to identify the subsets of genes which are consistently co-expressed in all of the provided datasets in a tuneable manner. However, results validation and parameter setting are issues that complicate the design of such methods. Moreover, although it is a common practice to test methods by application to synthetic datasets, the mathematical models used to synthesise such datasets are usually based on approximations which may not always be sufficiently representative of real datasets. RESULTS: Here, we propose an unsupervised method for the unification of clustering results from multiple datasets using external specifications (UNCLES). This method has the ability to identify the subsets of genes consistently co-expressed in a subset of datasets while being poorly co-expressed in another subset of datasets, and to identify the subsets of genes consistently co-expressed in all given datasets. We also propose the M-N scatter plots validation technique and adopt it to set the parameters of UNCLES, such as the number of clusters, automatically. Additionally, we propose an approach for the synthesis of gene expression datasets using real data profiles in a way which combines the ground-truth-knowledge of synthetic data and the realistic expression values of real data, and therefore overcomes the problem of faithfulness of synthetic expression data modelling. By application to those datasets, we validate UNCLES while comparing it with other conventional clustering methods, and of particular relevance, biclustering methods. We further validate UNCLES by application to a set of 14 real genome-wide yeast datasets as it produces focused clusters that conform well to known biological facts. Furthermore, in-silico-based hypotheses regarding the function of a few previously unknown genes in those focused clusters are drawn. CONCLUSIONS: The UNCLES method, the M-N scatter plots technique, and the expression data synthesis approach will have wide application for the comprehensive analysis of genomic and other sources of multiple complex biological datasets. Moreover, the derived in-silico-based biological hypotheses represent subjects for future functional studies.

Multimodulus algorithms (MMA) based adaptive blind equalizers mitigate inter-symbol interference in a digital communication system by minimizing dispersion in the quadrature components of the equalized sequence in a decoupled manner, i.e., the in-phase and quadrature components of the equalized sequence are used to minimize dispersion in the respective components of the received signal. These unsupervised equalizers are mostly incorporated in bandwidth-efficient digital receivers (wired, wireless or optical) which rely on quadrature amplitude modulation based signaling. These equalizers are equipped with nonlinear error-functions in their update expressions which makes it a challenging task to evaluate analytically their steady-state performance. However, exploiting variance relation theorem, researchers have recently been able to report approximate expressions for steady-state excess mean square error (EMSE) of such equalizers for noiseless but interfering environment. In this work, in contrast to existing results, we present exact steady-state tracking analysis of two multimodulus equalizers in a non-stationary environment. Specifically, we evaluate expressions for steady-state EMSE of two equalizers, namely the MMA2-2 and the βMMA. The accuracy of the derived analytical results is validated using different set experiments and found in close agreement.

The recently proposed UNCLES method has the ability to unify clustering results from multiple datasets under different types of external specifications. It can also tunably tighten the results such that many objects are unassigned from all of the clusters to obtain few tight clusters. Despite the success of this method, setting its parameters, such as the number of clusters (K) and the tuning parameters δ and (δ+, δ-), has never been automated. As its clusters vary in size, they cannot be validated by the existing validation indices. In this study we present a technique of validation based on our proposed M-N scatter plots. This technique has the ability to provide better fitness values for the clusters which include more objects while preserving their tightness. This well suits the nature of the results of UNCLES. We have applied this technique to a set of bacterial microarray datasets as well as a set of English vowels datasets. Our results demonstrate the success of the M-N plots in selecting the best few clusters out of a pool of clusters generated under varying K, δ, and (δ+, δ-) values. Our results also show that the best few clusters can be originated from different partitions, which shows the power of our technique in evaluating individual clusters rather than whole partitions. Finally, despite proposing this technique within the context of the UNCLES framework, it is readily applicable to other clustering results, especially when the parameters are not confidently predefined.

To meet the demand of clustering high dimensional data efficiently, in this paper, we propose a component-wise expectation conditional maximisation (CW-ECM) algorithm and integrate it within the recent proposed splitting-while-merging framework, which is called splitting-merging awareness tactics (SMART), for the mixture of factor analysers (MFA) model. The new algorithm has two advantages: it has ability to converge to actual or close actual number of clusters by a splitting-while-merging strategy, and it avoids the local maxima effectively and efficiently. Furthermore, we improve the splitting strategy in the original SMART framework and save more computational effort. We test out algorithm in two benchmark datasets and compare it with the state-of-the-art algorithms using many validation metrics. The results show that the proposed algorithm outperforms the compared algorithms in clustering performance with significantly less computational complexity.

BACKGROUND: The scale and complexity of genomic data lend themselves to analysis using sophisticated mathematical techniques to yield information that can generate new hypotheses and so guide further experimental investigations. An ensemble clustering method has the ability to perform consensus clustering over the same set of genes from different microarray datasets by combining results from different clustering methods into a single consensus result. RESULTS: In this paper we have performed comprehensive analysis of forty yeast microarray datasets. One recently described Bi-CoPaM method can analyse expressions of the same set of genes from various microarray datasets while using different clustering methods, and then combine these results into a single consensus result whose clusters' tightness is tunable from tight, specific clusters to wide, overlapping clusters. This has been adopted in a novel way over genome-wide data from forty yeast microarray datasets to discover two clusters of genes that are consistently co-expressed over all of these datasets from different biological contexts and various experimental conditions. Most strikingly, average expression profiles of those clusters are consistently negatively correlated in all of the forty datasets while neither profile leads or lags the other. CONCLUSIONS: The first cluster is enriched with ribosomal biogenesis genes. The biological processes of most of the genes in the second cluster are either unknown or apparently unrelated although they show high connectivity in protein-protein and genetic interaction networks. Therefore, it is possible that this mostly uncharacterised cluster and the ribosomal biogenesis cluster are transcriptionally oppositely regulated by some common machinery. Moreover, we anticipate that the genes included in this previously unknown cluster participate in generic, in contrast to specific, stress response processes. These novel findings illuminate coordinated gene expression in yeast and suggest several hypotheses for future experimental functional work. Additionally, we have demonstrated the usefulness of the Bi-CoPaM-based approach, which may be helpful for the analysis of other groups of (microarray) datasets from other species and systems for the exploration of global genetic co-expression.

Non-negative tensor factorization (NTF) has been successfully applied to analyze event-related potentials (ERPs), and shown superiority in terms of capturing multi-domain features. However, the time-frequency representation of ERPs by higher-order tensors are usually large-scale, which prevents the popularity of most tensor factorization algorithms. To overcome this issue, we introduce a non-negative canonical polyadic decomposition (NCPD) based on low-rank approximation (LRA) and hierarchical alternating least square (HALS) techniques. We applied NCPD (LRAHALS and benchmark HALS) and CPD to extract multi-domain features of a visual ERP. The features and components extracted by LRAHALS NCPD and HALS NCPD were very similar, but LRAHALS NCPD was 70 times faster than HALS NCPD. Moreover, the desired multi-domain feature of the ERP by NCPD showed a significant group difference (control versus depressed participants) and a difference in emotion processing (fearful versus happy faces). This was more satisfactory than that by CPD, which revealed only a group difference.

In this paper, we design and analyze a Newton-like blind equalization algorithm for the APSK system. Specifically, we exploit the principle of minimum entropy deconvolution and derive a blind equalization cost function for APSK signals and optimize it using Newton׳s method. We study and evaluate the steady-state excess mean square error performance of the proposed algorithm using the concept of energy conservation. Numerical results depict a significant performance enhancement for the proposed scheme over well established blind equalization algorithms. Further, the analytical excess mean square error of the proposed algorithm is verified with computer simulations and is found to be in good conformation.

Validity indices have been investigated for decades. However, since there is no study of noise-resistance performance of these indices in the literature, there is no guideline for determining the best clustering in noisy data sets, especially microarray data sets. In this paper, we propose a generalized parametric validity (GPV) index which employs two tunable parameters α and β to control the proportions of objects being considered to calculate the dissimilarities. The greatest advantage of the proposed GPV index is its noise-resistance ability, which results from the flexibility of tuning the parameters. Several rules are set to guide the selection of parameter values. To illustrate the noise-resistance performance of the proposed index, we evaluate the GPV index for assessing five clustering algorithms in two gene expression data simulation models with different noise levels and compare the ability of determining the number of clusters with eight existing indices. We also test the GPV in three groups of real gene expression data sets. The experimental results suggest that the proposed GPV index has superior noise-resistance ability and provides fairly accurate judgements.

In this paper, we propose a blind modulation classifier that differs from most existing classifiers. A low complexity minimum distance centroid estimator is suggested to estimate the channel gain and carrier phase jointly. The estimation is achieved by minimizing a signal-to-centroid distance. A new non-parametric likelihood function is proposed for fast classification with unknown noise variance and distribution. Numerical results show that the estimator provides reliable estimation of signal centroids, enabling an accurate classification with a non-parametric likelihood function. When different channel conditions are simulated, the proposed blind classifier achieves similar classification accuracy versus non-blind state-of-the-art classifiers while being more robust and having much lower complexity.

Many methods have been proposed to identify informative subsets of genes in microarray studies in order to focus the research. For instance, the recently proposed binarization of consensus partition matrices (Bi-CoPaM) method has, amongst its various features, the ability to generate tight clusters of genes while leaving many genes unassigned from all clusters. We propose exploiting this particular feature by applying the Bi-CoPaM over genome-wide microarray data from multiple datasets to generate more clusters than required. Then, these clusters are tightened so that most of their genes are left unassigned from all clusters, and most of the clusters are left totally empty. The tightened clusters, which are still not empty, include those genes that are consistently co-expressed in multiple datasets when examined by various clustering methods. An example of this is demonstrated in this paper for cyclic and acyclic genes as well as for genes that are highly expressed and that are not. Thus, the results of our proposed approach cannot be reproduced by other methods of genes' periodicity identification or by other methods of clustering.

In this paper, we propose a splitting-while-merging algorithm with finite mixture models (FMM) built on an improved splitting merging awareness tactics (SMART). The main property of SMART is that it does not require any dataset-dependent parameters or a priori knowledge about the datasets. The improved SMART framework integrates clustering selection criterion, which plays a vital role in the new algorithm. In the SMART-FMM implementation, the modified component-wise EM of mixtures is employed as a learning and merging technique and a model order selection algorithm is used as a clustering selection criterion. One demonstration example and one real microarray gene expression dataset are studied using our approach. The numerical results show that SMART-FMM is superior and more effective than others.

This study presents a method to analyze blood-oxygen-level-dependent (BOLD) functional magnetic resonance imaging (fMRI) signals associated with listening to continuous music. Semi-blind independent component analysis (ICA) was applied to decompose the fMRI data to source level activation maps and their respective temporal courses. The unmixing matrix in the source separation process of ICA was constrained by a variety of acoustic features derived from the piece of music used as the stimulus in the experiment. This allowed more stable estimation and extraction of more activation maps of interest compared to conventional ICA methods.

This study proposes a novel approach for the analysis of brain responses in the modality of ongoing EEG elicited by the naturalistic and continuous music stimulus. The 512-second long EEG data (recorded with 64 electrodes) are first decomposed into 64 components by independent component analysis (ICA) for each participant. Then, the spatial maps showing dipolar brain activity are selected in terms of the residual dipole variance through a single dipole model in brain imaging, and clustered into a pre-defined number (estimated by the minimum description length) of clusters. Subsequently, the temporal courses of the EEG theta and alpha oscillations of each component for each cluster are produced and correlated with the temporal courses of tonal and rhythmic features of the music. Using this approach, we found that the extracted temporal courses of the theta and alpha oscillations along central and occipital area of scalp in two of the selected clusters significantly correlated with the musical features representing progressions in the rhythmic content of the stimulus. We suggest that this demonstrates that with the proposed approach, we have managed to discover what kinds of brain responses were elicited when a participant was listening continuously to the long piece of naturalistic music.

This paper presents a new fragile watermarking method, whereby a secure block-wise and an interlaced watermarking mechanisms are hierarchically structured to provide higher tampering localisation capabilities. The block-wise method provides security against conventional distortions, cropping, as well as sophisticated attacks, whereas the interlaced scheme is aimed at enhancing the localisation accuracy achieved by the block-wise method. Results are presented to illustrate the improved localisation capabilities of the proposed method in comparison with some five existing watermarking schemes.

This paper presents a new fragile watermarking method, whereby two mechanisms are hierarchically structured to provide self-recovery capabilities. The first one is a secure block-wise mechanism, resilient to cropping, aimed at localising altered pixel-blocks. The second one is an iterative mechanism capable of reconstructing the original contents, by means of exhaustive attempts. The key features of the proposed method, which compare favourably to those of existing schemes, are low embedding distortion and resilience to cropping. Results demonstrate that the proposed scheme is capable of restoring the altered contents, even when the tampered region covers up to 32% of the total pixels in the image.

In this work, we derive three novel composite field arithmetic (CFA) Advanced Encryption Standard (AES) S-boxes of the field GF(((22)2)2). The best construction is selected after a sequence of algorithmic and architectural optimization processes. Furthermore, for each composite field constructions, there exists eight possible isomorphic mappings. Therefore, after the exploitation of a new common subexpression elimination algorithm, the isomorphic mapping that results in the minimal implementation area cost is chosen. High throughput hardware implementations of our proposed CFA AES S-boxes are reported towards the end of this paper. Through the exploitation of both algebraic normal form and seven stages fine-grained pipelining, our best case achieves a throughput 3.49 Gbps on a Cyclone II EP2C5T144C6 field-programmable gate array.

Automatic Modulation Classification (AMC) is an intermediate step between signal detection and demodulation. It is a very important process for a receiver that has no, or limited, knowledge of received signals. It is important for many areas such as spectrum management, interference identification and for various other civilian and military applications. This paper explores the use of Genetic Programming (GP) in combination with K-nearest neighbor (KNN) for AMC. KNN has been used to evaluate fitness of GP individuals during the training phase. Additionally, in the testing phase, KNN has been used for deducing the classification performance of the best individual produced by GP. Four modulation types are considered here: BPSK, QPSK, QAM16 and QAM64. Cumulants have been used as input features for GP. The classification process has been divided into two-stages for improving the classification accuracy. Simulation results demonstrate that the proposed method provides better classification performance compared to other recent methods.

A commonly considered image manipulation is to conceal undesirable objects or people in the scene with a region of pixels copied from the same image. Forensic mechanisms aimed at detecting this type of forgeries must also consider other potential types of post-processing, including geometric distortions. In this paper, a new method is proposed to detect duplicated regions, even when the cloned region has undergone reflection, rotation and scaling. The algorithm uses colour-dependent feature vectors to reduce the number of comparisons in the search stage, and one-dimensional (1-D) descriptors, invariant to reflection and rotation, to perform an efficient search in terms of memory usage. Comparison results are presented to evaluate the effectiveness of the proposed method and two existing schemes.

In this work, a genetic algorithm (GA) based frequency domain equalization (FDE) scheme referred to as FDE-GA, which does not require any guard interval (GI),is proposed for direct sequence-ultra wideband (DS-UWB) wireless communication systems and is shown to significantly outperform the RAKE receiver. The proposed FDE-GA receiver also has a dramatic complexity reduction over the previous RAKE-GA receiver, while achieving a comparable bit error rate (BER) performance. The FDE-GA structure achieves a much higher bandwidth efficiency than conventional FDE methods, because the inter-block-interference (IBI), as a result of the absence of the GI, is removed effectively within each block before the GA.

Abstract: The success of automatic classification is intricately linked with an effective feature selection. Previous studies on the use of genetic programming (GP) to solve classification problems have highlighted its benefits, principally its inherent feature selection (a process that is often performed independent of a learning method). In this paper, the problem of classification is recast as a feature generation problem, where GP is used to evolve programs that allow non-linear combination of features to create superFeatures, from which classification tasks can be achieved fairly easily. In order to generate superFeatures robustly, the binary string fitness characterization along with the comparative partner selection strategy is introduced with the aim of promoting optimal convergence. The techniques introduced are applied to two illustrative problems first and then to the real-world problem of audio source classification, with competitive results.

This study investigates Kalman smoothing (KS)-based frequency-domain channel estimation for uplink multiple-input multiple-output (MIMO) orthogonal frequency division multiple access (OFDMA) systems with time-varying channels. The proposed KS channel estimation scheme significantly outperforms the recursive least squares (RLS) channel estimation in the high signal-to-noise ratio (SNR) range, because of more effective exploitation of the signal information. In addition, channel interpolation is employed to improve the channel estimation accuracy by exploiting the correlation between adjacent subcarriers. The proposed KS channel estimator can also achieve a bit error rate (BER) performance which is close to the case with perfect channel state information (CSI) with a training overhead of only 5%.

The authors investigate dynamic physical resource block (PRB) allocation for the uplink long-term evolution (LTE) system with single carrier-frequency division multiple access (SC-FDMA). Three dynamic PRB allocation algorithms are proposed, which are referred to as the maximum greedy (MG), mean enhanced greedy (MEG) and single mean enhanced greedy (SMEG) algorithms, respectively. Simulation results show that the proposed algorithms significantly outperform the previous two-dimensional (2-D) algorithm in terms of bit error rate (BER) and data rate fairness. The MEG algorithm is shown to provide a performance close to the Hungarian algorithm (optimal algorithm to maximise the SE) in terms of spectral efficiency (SE), while requiring a much lower computational complexity. SMEG further reduces the complexity of MEG with little performance degradation. Furthermore, the effects of imperfect channel estimation, root mean square (RMS) delay, Doppler spread and channel estimate feedback delay on performance are investigated.

This paper investigates carrier frequency offset (CFO) estimation and inphase/quadrature (I/Q) imbalance compensation in time-varying frequency-selective channels. We first propose a linear least squares (LLS) CFO estimation approach which has a lower complexity and a higher accuracy than the previous nonlinear CFO estimation methods. We then propose a Kalman filtering based I/Q imbalance compensation approach in the presence of CFO, which demonstrates a good ability to track the channel time variations with a fast convergence speed, by nulling the cyclic prefix (CP) and including the CFO in the state vector of the equivalent channel model. The proposed Kalman filtering based I/Q imbalance compensation approach with associated equalization tracks the time variation with a fast convergence speed. Simulation results show that the proposed compensation approach for CFO and I/Q imbalance provides a bit error rate (BER) performance close to the ideal case with perfect channel state information (CSI), no CFO and no I/Q imbalance.

In some applications, such as surveillance cameras, it is essential that images acquired with the same device be deemed distinct. To fulfil this requirement, many fragile watermarking methods need to employ a different key for every single image. Keeping track of the correct key associated with each image may become a challenging task as the number of images increases. An existing image index-based scheme as a suitable solution for this problem is revisited, and its security limitations in applications are highlighted where higher localisation accuracy is required. Then, a new fragile watermarking scheme that enhances the localisation accuracy is proposed, while hindering brute force attacks. Experiments demonstrate that the proposed scheme outperforms the state-of-the-art private fragile watermarking methods.

A constant modulus algorithm is presented for blind equalization of complex-valued communication channels. The proposed algorithm is obtained by solving a novel deterministic optimization criterion which comprises the minimization of a priori as well as a posteriori dispersion error, leading to an update equation having a particular zero-memory continuous Bussgang-type nonlinearity. We also derive a stochastic bound for the range of step-sizes for a generic Bussgang-type constant modulus algorithm. The theoretical result is validated through computer simulations.

In this letter, we adaptively optimize the equalizer output energy to obtain a joint blind equalization and carrier-phase recovery solution. The resulting (multimodulus) update algorithm possesses a particular zero-memory Bussgang-type nonlinearity. We provide evidence of good performance, in comparison to existing adaptive methods, like RCA, MMA and CMA, through computer simulations for higher-order quadrature amplitude modulation signalling on symbol- and fractionally-spaced channels.

By generalizing and modifying some existing cost-functions, we present two new, generic and efficient multimodulus families of blind equalization algorithms for use in higher-order quadrature amplitude modulation based digital communication systems. Proposed algorithms are shown to be capable of blindly equalizing and recovering carrier-phase at convergence speed much faster than existing counterparts on certain QAM sizes. We show that particular examples of the proposed cost-functions include a number of existing algorithms. We also provide detailed dynamic convergence analysis which is found in good conformation with those obtained from Monte-Carlo experiments.

We propose a novel semi-blind compensation scheme for both frequency-dependent and frequencyindependent I/Q imbalance based on independent component analysis (ICA) in multiple-input multiple-output (MIMO) orthogonal frequency division multiplexing (OFDM) systems, where ICA is applied to compensate for I/Q imbalance and equalize the received signal jointly, without any spectral overhead. A reference signal is embedded in the transmitted signal with little power consumption and no spectral overhead introduced, to enable ambiguity elimination for the ICA output signal at the receiver. Moreover, channel interpolation is incorporated with layered space frequency equalization (LSFE) to enhance the system performance. Simulation results show that the proposed implicit compensation scheme can not only provide a better bit error rate (BER) performance and a higher bandwidth efficiency than the previous training based I/Q imbalance compensation method, but also outperform the ideal case with perfect channel state information (CSI) and no I/Q imbalance, due to additional frequency diversity.

I/Q imbalance and carrier frequency offset (CFO) are two typical radio frequency (RF) circuit analog impairments in wireless communication systems, and degrade the system performance severely. In this paper, we propose a novel Kalman filtering based compensation scheme for I/Q imbalance and CFO in time-varying MIMO OFDM systems. To circumvent CFO and track time variations of wireless communication channels, the CFO is absorbed into the state vector of the equivalent channel model. Moreover, the inter-carrier interference (ICI) caused by phase shift of the cyclic prefix (CP) in the equivalent system is removed by decision feedback filtering, which allows low complexity compensation on each subcarrier independently. Simulation results show that the proposed approach can compensate for I/Q imbalance and CFO effectively and is robust again time variations.

I/Q imbalance, which is one of the radio frequency (RF) circuit impairments in direct conversion transmitter and receiver, introduces severe performance degradation in wireless communication systems. In this paper, we propose a novel blind compensation algorithm for both frequency-dependent and frequency-independent I/Q imbalance based on independent component analysis (ICA) in multiple input multiple output (MIMO) orthogonal frequency division multiplexing (OFDM) systems, where ICA, an efficient higher order statistics (HOS) based blind source separation technique, is applied to compensate for I/Q imbalance and equalize the received signals simultaneously. Moreover, preceding is employed to resolve the ambiguity in the ICA output signals. Simulation results show that the proposed approach can not only compensate for I/Q imbalance effectively, but also achieve frequency diversity gains and outperform the case with perfect channel state information (CSI) and no I/Q imbalance.

In this paper, we propose an enhanced greedy dynamic subcarrier allocation algorithm for single carrier frequency division multiple access (SC-FDMA) systems. This so called mean-enhanced greedy algorithm allocates subcarriers in a greedy fashion, based on the information of the users' subcarriers mean gains. We show through simulation results that the proposed algorithm outperforms the conventional greedy algorithm. It also outperforms the benchmark Hungarian algorithms in terms of bit error rate (BER) with a lower computational complexity. Furthermore, the proposed algorithm is generic and can be easily extended for orthogonal frequency division multiple access (OFDMA). We compare the performance of SC-FDMA and OFDMA, and point out that for a high number of users, there is no frequency diversity gain of SC-FDMA over OFDMA as both systems demonstrate benefit from multiuser diversity.

We propose a genetic algorithm (GA) based equalization approach for direct sequence ultra-wideband (DS-UWB) wireless communications, where GA is combined with a RAKE receiver to combat the inter-symbol interference (ISI) due to the frequency selective nature of UWB channels for high data rate transmission. Simulation results show that the proposed GA based structure significantly outperforms the RAKE receiver. It also provides a close bit error rate (BER) performance to the optimal maximum likelihood detection (MLD) approach, while requiring a much lower computational complexity.

Abstract: We show that tuning of multiple parameters for a 2‐norm support vector machine (L2‐SVM) could be viewed as an identification problem of a nonlinear dynamic system. Benefiting from the reachable smooth nonlinearity of an L2‐SVM, we propose to employ the extended Kalman filter to tune the kernel and regularization parameters automatically for the L2‐SVM. The proposed method is validated using three public benchmark data sets and compared with the gradient descent approach as well as the genetic algorithm in measures of classification accuracy and computing time. Experimental results demonstrate the effectiveness of the proposed method in higher classification accuracies, faster training speed and less sensitivity to the initial settings.

Abstract: In this paper, a partial supervision strategy for a recently developed clustering algorithm, the nearest neighbour clustering algorithm (NNCA), is proposed. The proposed method (NNCA‐PS) offers classification capability with a smaller amount of a priori knowledge, where a small number of data objects from the entire data set are used as labelled objects to guide the clustering process towards a better search space. Experimental results show that NNCA‐PS gives promising results of 89% sensitivity at 95% specificity when used to segment retinal blood vessels, and a maximum classification accuracy of 99.5% with 97.2% average accuracy when applied to a breast cancer data set. Comparisons with other methods indicate the robustness of the proposed method in classification. Additionally, experiments on parallel environments indicate the suitability and scalability of NNCA‐PS in handling larger data sets.

We propose two variations of the support vector data description (SVDD) with negative samples (NSVDD) that learn a closed spherically shaped boundary around a set of samples in the target class by involving different forms of slack vectors, including the two-norm NSVDD and nu-NSVDD. We extend the NSVDDs to solve the multiclass classification problems based on the distances between the samples and the centers of the learned spherically shaped boundaries in a kernel-defined feature space by using a combination of linear discriminant analysis (LDA) and nearest-neighbor (NN) rule. Extensive simulations are developed with one real-world data set on the automatic monitoring of roller bearings with vibration signals and eight benchmark data sets for both binary and multiclass classification. The benchmark testing results show that our proposed methods provide lower classification error rates and smaller standard deviations with the cross-validation procedure. The two-norm NSVDD with the LDA-NN rule recorded a test accuracy of 100.0% for the binary fault detection of roller bearings and 99.9% for the multiclass classification of roller bearings under six conditions.

We propose a non-redundant linear precoding scheme and three peak-to-average power ratio (PAPR) reduction schemes for multiple-input multiple-output (MIMO) orthogonal frequency division multiplexing (OFDM) systems with independent component analysis (ICA) based blind equalization. The proposed precoding at the transmitter allows complete elimination of the ambiguity in the ICA equalized signals under certain conditions. The optimal design of precoding is investigated, and performance analysis on the ambiguity error probability is provided. The proposed PAPR reduction schemes are incorporated with precoding, and therefore do not introduce any spectral overhead compared to conventional PAPR reduction schemes. Simulation results show that the proposed blind structure provides a bit error rate (BER) performance which is much better than that of the subspace method, and close to the case with perfect channel state information (CSI) at the receiver. Furthermore, the proposed structure can reduce the PAPR of the transmit signals considerably.

A new fragile logo watermarking scheme is proposed for public authentication and integrity verification of images. The security of the proposed block-wise scheme relies on a public encryption algorithm and a hash function. The encoding and decoding methods can provide public detection capabilities even in the absence of the image indices and the original logos. Furthermore, the detector automatically authenticates input images and extracts possible multiple logos and image indices, which can be used not only to localise tampered regions, but also to identify the original source of images used to generate counterfeit images. Results are reported to illustrate the effectiveness of the proposed method.

Four new features for the analysis of breast masses are presented. These features were designed to be insensitive to the exact shape of the contour of the masses, so that an approximate contour, such as one extracted via an automated segmentation algorithm, can be employed in their computation. The features measure the degree of spiculation of a mass and the local fuzziness of the mass margins. The features were tested for characterization (discrimination between circumscribed and spiculated) and diagnosis (discrimination between benign and malignant) of breast masses, using 319 masses and three different classifiers. Approximately 90% and 76% of correct classification in characterization and diagnosis, respectively, were achieved.

We propose an adaptive Turbo multiuser detection and co-channel interference (CCI) suppression scheme for the uplink single-carrier frequency division multiple access (SC-FDMA) system, with multiple antennas employed for both each user and the base station. Turbo detection is incorporated with the simplified Turbo recursive- least-square (RLS) channel estimator, which provides close performance to the case with perfect channel state information (CSI). To suppress unknown CCI, we design a block-wise estimator, based on low pass smoothing, to estimate the spatial correlation matrix of unknown CCI plus noise.

Soft decision directed Turbo frequency-domain channel estimation is investigated for single-carrier (SC) multiuser detection. A Kalman filtering based channel estimation algorithm is first proposed under an uncorrelated- channel-tap state space model, which is the optimal linear unbiased channel estimator for the case of interest. Then, a normalized Turbo recursive least-square (RLS) based channel estimation algorithm is proposed for a more realistic channel model. Simulation results show that both the proposed channel estimation algorithms provide a close performance to the case with perfect channel state information (CSI), and outperform the previously proposed Turbo frequency-domain channel estimation in fast fading channels.

Abstract: Feature extraction helps to maximize the useful information within a feature vector, by reducing the dimensionality and making the classification effective and simple. In this paper, a novel feature extraction method is proposed: genetic programming (GP) is used to discover features, while the Fisher criterion is employed to assign fitness values. This produces non‐linear features for both two‐class and multiclass recognition, reflecting the discriminating information between classes. Compared with other GP‐based methods which need to generate c discriminant functions for solving c‐class (c>2) pattern recognition problems, only one single feature, obtained by a single GP run, appears to be highly satisfactory in this approach. The proposed method is experimentally compared with some non‐linear feature extraction methods, such as kernel generalized discriminant analysis and kernel principal component analysis. Results demonstrate the capability of the proposed approach to transform information from the high‐dimensional feature space into a single‐dimensional space by automatically discovering the relationships between data, producing improved performance.

Single carrier (SC) code division multiple access (CDMA) with block transmission has been shown to be more effective while utilising a low-complexity equaliser to combat frequency-selective fading channels, when compared with conventional direct sequence CDMA technology. It also has lower peak-to-average power ratio and lower frequency sensitivity compared with multicarrier CDMA. The authors propose two blind channel estimation methods for uplink multiple input multiple output SC-CDMA systems with block transmisssion-one is the subspace-based method and the other is the so-called autocorrelation contribution method (ACM). Both the methods provide close performance to the case with perfect channel knowledge at high signal-to-noise ratio (SNR) without any training data required. It is shown that ACM yields a better performance than the subspace method at a lower SNR, and a similar performance at a high SNR, with the advantages of avoiding rank determination and noise power estimation as in the subspace method. In addition, the authors integrate layered space frequency equalisation with blind channel estimation, which provides improved performance over the conventional linear equalisation, by employing successive interference cancellation.

The premise behind all evolutionary methods is ldquosurvival of the fittest,rdquo and consequently, individuals require a quantitative fitness measure. This paper proposes a novel strategy for evaluating individual's relative strengths and weaknesses, as well as representing these in the form of a binary string fitness characterization (BSFC); in addition, as customary, an overall fitness value is assigned to each individual. Utilizing the BSFC, we demonstrate both novel population evaluation measures and a pairwise mating strategy, comparative partner selection (CPS), with the aim of evolving a population that promotes effective solutions by reducing population-wide weaknesses. This strategy is tested with six standard genetic programming benchmarking problems.

This letter proposes a novel semi-blind layered space-frequency equalization (LSFE) receiver for single- carrier cyclic-prefix (SC-CP) multiple-input multiple-output (MIMO) systems with block transmission, based on independent component analysis (ICA). Simulation results show that semi- blind LSFE with various Doppler shifts can provide performance close to the case with perfect channel state information (CSI), using a training overhead of only 0.05%. It also significantly outperforms its orthogonal frequency division multiplexing (OFDM) counterpart with perfect CSI and semi-blind linear SC frequency domain equalization (SC-FDE) over a wide range of signal to noise ratios (SNRs).

By combining single-carrier (SC) frequency domain equalization (FDE) and Turbo equalization, we propose a minimum mean square error (MMSE) based block-wise low complexity Turbo space-frequency equalization (TSFE) structure, which offers a tremendous complexity reduction over the symbol-wise TSFE and Turbo time-domain equalization (TTDE) structures. With a moderate code rate, the proposed SC TSFE significantly outperforms its Turbo OFDM (TOFDM) counterpart, at a comparable complexity. Besides, TSFE based on the least-mean-square structured channel estimation (LMS-SCE) provides a performance close to the case with perfect channel state information (CSI), at a high convergence rate.

Adaptive turbo frequency-domain channel estimation is incorporated with low complexity turbo space-frequency equalization (TSFE) for single-carrier (SC) multi-user detection. The simplified turbo recursive least square (RLS) channel estimation algorithm provides nearly the same performance as its full complexity version, with a tremendous complexity reduction. With PSK modulations, the simplified turbo RLS channel estimation reduces to turbo LMS channel estimation. With a low training overhead, the simplified turbo RLS channel estimation provides a performance comparable to the case with perfect channel state information (CSI).

A method for detection of masses in mammograms is presented. This method follows the general scheme of: (1) preprocessing of the image to increase the signal-to-noise ratio of the lesions being detected, (2) segmentation of all potential lesions, and (3) elimination of false-positive findings. An algorithm for enhancement of mammograms is proposed which has the objective of improving the segmentation of distinct structures in mammograms. The enhancement algorithm uses wavelet decomposition and reconstruction, morphological operations, and local scaling. After preprocessing, the segmentation of regions is performed via conversion to binary images at multiple threshold levels, and a set of features is computed from each of the segmented regions. A ranking system based on the features computed is also presented. This system is employed to select the regions representing abnormalities. The method was tested on 57 mammographic images of masses from the mini-MIAS database, including circumscribed, spiculated, and ill-defined masses. In this test, the proposed method achieved a sensitivity of 80% at 2.3 false-positives (FPs) per image.

We propose a 2-plane learning method for binary classification, named as the strict 2-surface proximal (S2SP) classifier, by seeking two cross proximal planes based on two strict optimization objectives with a "square of sum" optimization factor, of which the nonlinearity is achieved by employing kernel functions. We apply the S2SP classifier for both linear and nonlinear classification to recognize malignant tumors from a set of 57 regions in mammograms, of which 20 are related to malignant tumors and 37 to benign masses. Ten different feature combinations are studied. Experimental results demonstrate that the linear S2SP classifier provides results comparable to those obtained by Fisher linear discriminant analysis (FLDA). For one feature set (FSs), the linear classification performance was significantly improved to 0.97 by using the S2SP classifier, as compared to the FLDA performance of 0.82, in terms of the area under the receiver operating characteristics (ROC) curve. In the case of nonlinear classification, the S2SP classifier with the triangle kernel provided a perfect performance of 1.0 for all of the ten feature combinations, also evaluated in terms of the area under the ROC curve, but with good robustness limited to the setting of the kernel parameter in a certain range.

Adaptive Turbo frequency-domain channel estimation is investigated for single-carrier (SC) multi-user detection in the presence of unknown co-channel interference (CCI). We propose a modified Turbo recursive least square (RLS) channel estimation algorithm which provides very close performance to Turbo RLS channel estimation, with a huge complexity reduction. It also significantly outperforms the Turbo least mean square (LMS) channel estimation in terms of performance and convergence speed, and requires a similar complexity to the LMS channel estimation in the scenario of phase shift keying (PSK) modulation. Beside, we incorporate adaptive Turbo channel estimation with low complexity block wise Turbo Space-frequency equalization and CCI suppression (TSFE-CCIS), which also operates in the frequency domain.

In this paper, we propose the application of two blind channel estimation algorithms namely the subspace approach and the autocorrelation contribution method (ACM) algorithm for multiple input multiple output (MIMO) single-carrier CDMA (SC-CDMA) block transmission systems in the quasi- synchronous uplink scenario. Using only second order statistics, the ACM approach shows similar performance to that of subspace based approach with the added advantage of eliminating the need for rank estimation and noise power calculation as in subspace technique. The guard band utilised serves the dual role of eliminating ISI (inter-signal interference) and aiding the blind channel estimation stage. The use of cyclic prefix (CP) as well as zero padded (ZP) guard band for blind channel estimation is explained with simulation results highlighting the robust nature of ZP based schemes to blind channel estimation as compared to CP based system.

Robust automatic speaker verification has become increasingly desirable in recent years with the growing trend toward remote security verification procedures for telephone banking, bio-metric security measures and similar applications. While many approaches have been applied to this problem, genetic programming offers inherent feature selection and solutions that can be meaningfully analyzed, making it well suited to this task. This paper introduces a genetic programming system to evolve programs capable of speaker verification and evaluates its performance with the publicly available TIMIT corpora. We also show the effect of a simulated telephone network on classification results which highlights the principal advantage, namely robustness to both additive and convolutive noise

We propose a novel blind orthogonal frequency division multiplexing (OFDM) receiver structure for multiple- input multiple-output (MIMO) systems based on independent component analysis (ICA), which increases the spectral efficiency effectively compared to training based systems, and provides considerable performance enhancement over previous ICA based methods. To further improve the performance, we also incorporate ICA with iterative layered space-time equalization (LSTE), which provides performance close to the case with perfect channel state information (CSI). The reduced-complexity versions of the two proposed structures, which use channel interpolation, result in similar performance at a substantially lower computational cost. Our receiver structures also have significantly better performance and lower complexity than a previously proposed subspace method when a large number of subcarriers are employed.

In this contribution, we propose an entirely novel family of flexible score functions for blind source separation (BSS), based on the generalized Gamma family of densities. An efficient maximum likelihood (ML) technique for estimating the parameters of such score functions in an adaptive BSS setup, is also put forward. Simulations indicate that the proposed density model can approximate speech signals more accurately than conventional distributions, which leads to an increase in separation performance and convergence speed

In this paper, a new feature vector for each pixel, in conjunction with the K-nearest neighbour classifier, is proposed for the segmentation of retinal blood vessels in digital colour fundus images. The proposed feature vector consists of two scale-space features - the largest eigenvalue and the gradient magnitude - of the intensity image, representing the two attributes of any vessel, i.e. the piecewise linearity and parallel edges, as well as the green channel image intensity. In terms of sensitivity and specificity, our results are comparable with other supervised method which uses a set of 31 features, yet in terms of processing time, our method uses a smaller number of features and results in a significant reduction in the processing time

This paper addresses the blind separation of convolutive and temporally correlated mixtures of speech, through the use of a multichannel blind deconvolution (MBD) method. In the proposed framework (LP-NGA), spatio-temporal separation is carried out by entropy maximization using the well-known natural gradient algorithm (NGA), while a temporal pre-whitening stage, based on linear prediction (LP), manages to fully preserve the original spectral characteristics of each source contribution. Confronted with synthetic convolutive mixtures, we show that the LP-NGA-an unconstrained natural extension to the multichannel BSS problem-benefits not only from fewer model constraints, but also from other factors, such as an overall increase in separation performance, spectral preservation efficiency and speed of convergence.

A number of commonly used methods for estimating the exponent parameter of a generalized Gaussian density (GGD) are reviewed, described and compared. More importantly, focusing on the family of entropy matching estimators (EMEs), a novel entropic expression with respect to higher-order moments of the modelled data is proposed. This yields an elegant generalized entropy matching estimator (G-EME). Comparative experimental results illustrate the high accuracy of the proposed estimator, for both light- and heavy-tailed distributions, as well as speech data.

Genetic programming has seldom been used for multi-class classification purposes. Previously, it was achieved by separating the output manually for different classes or expanding an n-class problem to n two-class problems. In this paper, we present a new approach to solving multi-class problems by using genetic programming for feature generation, and applying the K-nearest neighbor as a classifier. The results are comparable with other classifiers.

This paper proposes a reduced complexity blind layered space-time equalization (LSTE) for multiple-input multiple-output (MIMO) orthogonal frequency division multiplexing (OFDM). Blind source separation (BSS) combined with a vertical Bell Laboratories layered space-time (V-BLAST) scheme is used in a small portion of subcarriers to perform blind signal detection and channel estimation iteratively. In the remaining subcarriers the channel state information (CSI) is obtained by interpolation using the previously acquired CSI, and signal detection is subsequently carried out with a V-BLAST scheme. Compared to previous work where BSS is used in each subcarrier, the proposed system obtains similar performance while significantly reducing the computational complexity and improving the robustness to the propagation of permutation errors. Also, the performance of the blind LSTE can approach the ideal case with perfect CSI, by using a moderate number of iterations.

One of the major challenges in pattern recognition problems is the feature extraction process which derives new features from existing features, or directly from raw data in order to reduce the cost of computation during the classification process, while improving classifier efficiency. Most current feature extraction techniques transform the original pattern vector into a new vector with increased discrimination capability but lower dimensionality. This is conducted within a predefined feature space, and thus, has limited searching power. Genetic programming (GP) can generate new features from the original dataset without prior knowledge of the probabilistic distribution. In this paper, a GP-based approach is developed for feature extraction from raw vibration data recorded from a rotating machine with six different conditions. The created features are then used as the inputs to a neural classifier for the identification of six bearing conditions. Experimental results demonstrate the ability of GP to discover autimatically the different bearing conditions using features expressed in the form of nonlinear functions. Furthermore, four sets of results--using GP extracted features with artificial neural networks (ANN) and support vector machines (SVM), as well as traditional features with ANN and SVM--have been obtained. This GP-based approach is used for bearing fault classification for the first time and exhibits superior searching power over other techniques. Additionaly, it significantly reduces the time for computation compared with genetic algorithm (GA), therefore, makes a more practical realization of the solution.

This paper addresses the blind separation of convolutive and temporally correlated speech mixtures, through the use of a multichannel blind deconvolution (MBD) method. In the proposed method (NGA-LP) spatio-temporal separation is achieved by entropy maximization using the natural gradient algorithm (NGA), while a temporal prewhitening stage, based on linear prediction (LP), preserves the original spectral characteristics of each source contribution. It is further shown that a parameterized optimal nonlinearity derived from the generalized Gaussian density (GGD) model, increases the overall separation performance. Experiments with convolutive mixtures illustrate the merits of the proposed method.

We propose a simple power allocation (PA) scheme for wireless MIMO systems with layered space-time equalization (LSTE), which aims at minimizing the overall BERs for all data streams. This so-called MIN-BER PA scheme significantly outperforms the uniform power (UP) case in a wide range of RMS delay spreads and at high SNR, with little loss in computational complexity and bandwidth efficiency. Compared to EQ-BER PA, MIN-BER PA yields close performance, but requires much less complexity especially at high SNR. Closed-form results are provided for MIN-BER PA, based on which intensive performance analysis is made, including BER performance and signal detection ordering for LSTE with PA.

A new unsupervised fuzzy reinforcement learning vector quantization (FRLVQ) algorithm for image compression based on the combination of fuzzy K-means clustering algorithm and topology knowledge is proposed. In each iteration of reinforcement learning (RL), the size and direction of the movement of a codevector is decided by the overall pair-wise competition between the attraction of each training vector and the repellent force of the corresponding winning codevector. While each training vector only affects the winning codevector in the generalised Lloyd algorithm (GLA) strategy, and only the attraction of training vectors are considered in the fuzzy K-means (FKM) strategy. The competition is measured by the membership function. Simulation results are presented to compare the proposed FRLVQ with GLA and FKM algorithms. It is apparent that FRLVQ has the better quality of codebook design, is very insensitive to the selection of the initial codebook, and relatively insensitive to the choice of learning rate sequences.

This contribution provides a unified framework for the analytic or closed-form estimators for blind separation of independent source signals in real-valued instantaneous linear mixtures, in the noiseless two-source two-sensor scenario. First, the connections among three existing 4th-order analytic formulae (CF, AML and EML) are clarified. Next, a general expression for the estimation of the relevant separation parameter from the data nth-order statistics is unveiled, of which the extended maximum likelihood (EML) estimator is a particular case at n=4, and from which a novel third-order estimator is derived at n=3. Asymptotic performance analysis results are also presented. Associated contrast-function optimization criteria are shown to extend the applicability of a known 4th-order contrast function in the two-signal case. Simulations illustrate and validate the theoretical exposition.

We present a computationally simple algorithm suitable for fast, high resolution estimation of time delays and Doppler shifts (which are necessary for target localization and tracking) between narrowband signals in an active sonar system. The algorithm uses a modulated Lagrange interpolation filter and an LMS-type algorithm. The problem of delay and Doppler estimation is reduced to a linear regression problem. Convergence and performance analysis of the method is studied both analytically and through simulation. It is demonstrated that the method provides estimates close to the Cramer-Rao lower bound.

The problem of estimating the phase parameters of a phase modulated signal in the presence of coloured multiplicative noise (random amplitude modulation) and additive white noise, both Gaussian, is addressed. Closed-form expressions for the exact and large-sample Cramer-Rao bounds (CRB) are derived. It is shown that the CRB is not significantly affected by the colour of the modulating process, especially when the signal-to-noise ratio is high. Hence, maximum likelihood type estimators which ignore the noise colour and optimize a criterion with respect to only the phase parameters are proposed. These estimators are shown to be equivalent to the nonlinear least squares estimators which consist of matching the squared observations with a constant amplitude phase modulated signal when the mean of the multiplicative noise is forced to zero. Closed-form expressions are derived for the efficiency of these estimators, and are verified via simulations.

Blind source separation aims to extract a set of independent signals from a set of observed linear mixtures. After whitening the sensor output, the separation is achieved by estimating an orthogonal transformation, which in the real-mixture two-source two-sensor case is entirely characterized by a single rotation angle. This contribution studies an estimator of such an angle. Even though it is derived from geometric notions based on the scatter-plots of the signals involved, it is found, empirically, to exhibit a performance clearly up to the mark of other methods based on optimality criteria and, theoretically, to improve and generalize one of such procedures. The simplicity of the suggested estimator results in a straightforward adaptive version, which converges regardless of the source distribution, for quite mild conditions, and whose asymptotic analysis is easy to carry out. The applicability of the estimator in a full separation system is also illustrated.

This letter addresses the problem of on-line sub-sample time delay estimation of narrowband signals of known center frequency. A new form of the Lagrange interpolator filter is presented in this letter, which is incorporated into the explicit time delay estimator (ETDE) method. Simulations show that ETDE with the filter modulated to the signal center frequency significantly outperforms conventional ETDE.

This paper addresses the problem of detecting deterministic signals in multiplicative noise. The multiplicative noise model is appropriate for modelling coherent imaging systems such as SAR and laser. Locally optimum (LO) detectors are derived for any arbitrary multiplicative noise distribution. The gamma and generalized Gaussian distributions are studied in detail. We also introduce an extension of the generalized Gaussian density to include asymmetry. The performance of the LO detectors is studied and compared with that of the linear correlation detector. The paper gives insight into the influence of the tail length of the noise distribution on the detection power.

This paper introduces two algorithms for analog and digital modulations recognition. The first algorithm utilizes the decision-theoretic approach in which a set of decision criteria for identifying different types of modulations is developed. In the second algorithm the artificial neural network (ANN) is used as a new approach for the modulation recognition process. Computer simulations of different types of band-limited analog and digitally modulated signals corrupted by band-limited Gaussian noise sequences have been carried out to measure the performance of the developed algorithms. In the decision-theoretic algorithm it is found that the overall success rate is over 94% at the signal-to-noise ratio (SNR) of 15 dB, while in the ANN algorithm the overall success rate is over 96% at the SNR of 15 dB.

Vibration analysis can give an indication of the condition of a rotating shaft highlighting potential faults such as unbalance and rubbing. Faults may however only occur intermittently and consequently to detect these requires continuous monitoring with real time analysis. This paper describes the use of artificial neural networks (ANNs) for classification of condition and compares these with other discriminant analysis methods. Moments calculated from time series are used as input features as they can be quickly computed from the measured data. Orthogonal vibrations are considered as a two-dimensional vector, the magnitude of which can be expressed as time series. Some simple signal processing operations are applied to the data to enhance the differences between signals and comparison is made with frequency domain analysis.

Several methods based on higher order statistics are currently used to solve the blind source separation (BSS) problem. In the present letter an approach from Mansour and Jutten (1995) is considered and simplified via the application of the singular value decomposition, which decorrelates the observed signals and leads to a straightforward solution based on some results developed by Comon (1989).

Various properties of wave breaking in the sea have been investigated by several authors, but none of these has proved to be very satisfactory in predicting wave breaking events. As this phenomenon is highly nonlinear, a fresh attempt is made to discriminate breaking waves, using real data from a wave tank, based primarily on the third-order statistics. It is shown that bispectral analysis offers a possible solution to the problem of predicting breaking wave events from a given time series.>