Search

Deep learning technology has been widely used in object detection. Although the deep learning technology greatly improves the accuracy of object detection, we also have the challenge of a high computational time. You Only Look Once (YOLO) is a network for object detection in images. In this paper, we propose a real-time object detection algorithm for videos based on the YOLO network. We eliminate the influence of the image background by image preprocessing, and then we train the Fast YOLO model for object detection to obtain the object information. Based on the Google Inception Net (GoogLeNet) architecture, we improve the YOLO network by using a small convolution operation to replace the original convolution operation, which can reduce the number of parameters and greatly shorten the time for object detection. Our Fast YOLO algorithm can be applied to real-time object detection in video. [ABSTRACT FROM AUTHOR]

Tongue diagnosis is one of the important methods in the Chinese traditional medicine. Doctors can judge the disease’s situation by observing patient’s tongue color and texture. This paper presents a novel approach to extract color and texture features of tongue images. First, we use improved GLA (Generalized Lloyd Algorithm) to extract the main color of tongue image. Considering that the color feature cannot fully express tongue image information, the paper analyzes tongue edge’s texture features and proposes an algorithm to extract them. Then, we integrate the two features in retrieval by different weight. Experimental results show that the proposed method can improve the detection rate of lesion in tongue image relative to single feature retrieval. [ABSTRACT FROM AUTHOR]

With the flooding datasets of medical Computed Tomography (CT) and Magnetic Resonance Imaging (MRI), implicit modeling techniques are increasingly applied to reconstruct the human organs, especially the vasculature. However, displaying implicitly represented geometric objects arises heavy computational burden. In this study, a Graphics Processing Unit (GPU) accelerating technique was developed for high performance rendering of implicitly represented objects, especially the vasculatures. The experimental results suggested that the rendering performance was greatly enhanced via exploiting the advantages of modern GPUs. [ABSTRACT FROM AUTHOR]

With the flooding datasets of medical Computed Tomography (CT) and Magnetic Resonance Imaging (MRI), implicit modeling techniques are increasingly applied to reconstruct the human organs, especially the vasculature. However, displaying implicitly represented geometric objects arises heavy computational burden. In this study, a Graphics Processing Unit (GPU) accelerating technique was developed for high performance rendering of implicitly represented objects, especially the vasculatures. The experimental results suggested that the rendering performance was greatly enhanced via exploiting the advantages of modern GPUs. [ABSTRACT FROM AUTHOR]

Abstract: Based on the beam coherent-polarization (BCP) matrix approach and propagation law of partially coherent beams, analytical propagation equations of partially polarized Gaussian Schell-model (PGSM) beams through an astigmatic lens are derived, which enables us to study the propagation-induced polarization changes and irradiance distributions at any propagation distance of PGSM beams through an astigmatic lens within the framework of the paraxial approximation. Detailed numerical results for a PGSM beam passing through an astigmatic lens are presented. A comparison with the aberration-free case is made, and shows that the astigmatism affects the propagation properties of PGSM beams. [Copyright &y& Elsevier]

Recently, a deep learning model, the deep forest (DF), was designed as an alternative to deep neural networks. Each cascade layer of the DF contains a set of random forests (RFs) with a large number of decision trees, some of which are of high redundancy and poor performance. To avoid the negative impacts of such decision trees, this paper proposes to optimize RFs in each cascade layer of the DF so as to realize a pruned deep forest (PDF) with higher performance and smaller ensemble size. In this paper, a new ordering-based ensemble pruning method is proposed based on feature vectorization and quantum walks. This method simultaneously considers the accuracy and the diversity of base classifiers, and it provides an integrated evaluation criterion for ordering base classifiers in the ensemble system. The effectiveness of the proposed method is verified by experiments and discussions. [ABSTRACT FROM AUTHOR]

Most deep learning methods have inherent defects and are rarely applied in the classification task of small-sized imbalanced datasets. On the one hand, data imbalance causes the classification results of the model to be biased toward the majority class. On the other hand, limited training data results in over-fitting. Deep forest (DF) is an interesting deep learning model that can perfectly work on small-sized datasets, and its performance is highly competitive with deep neural networks. In the present study, a variant of the DF called the imbalanced deep forest (IMDF) is proposed to effectively improve the classification performance of the minority class. It aims to explore the application of deep learning on small-sized imbalanced datasets. The IMDF is the cascade of multiple layers, where each layer is the ensemble of multiple units. The main idea behind the proposed method is to enable each unit of the IMDF to handle imbalanced data so that the classification results of the entire IMDF are biased toward minority class. Performed experiments demonstrate the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]

Partial label learning (PLL) is a class of weak supervision learning problems in which each data sample has a candidate set of labels, among which only one label is correct. In this paper, a new PLL algorithm with prior information of the label distribution based on ECOC (PL-PIE) is proposed. PL-PIE utilizes the ECOC framework to decompose the problem into multiple binary problems. Different from the instability of the existing random dichotomy, the proposal exploits the prior information of label distribution to generate positive and negative classes with stable performance. Extensive experimental results demonstrate that the proposed PL-PIE algorithm has highly competitive performance compared to the state-of-the-art PLL algorithms. [ABSTRACT FROM AUTHOR]

Effective representation of objects in irregular and unordered point clouds is one of the core challenges in 3D vision. Transforming point cloud into regular structures, such as 2D images and 3D voxels, are not ideal. It either obscures the inherent geometry information of 3D data or results in high computational complexity. Learning permutation invariance feature directly from raw 3D point clouds using deep neural network is a trend, such as PointNet and its variants, which are effective and computationally efficient. However, these methods are weak to reveal the spatial structure of 3D point clouds. Our method is delicately designed to capture both global and local spatial layout of point cloud by proposing a Local k-NNs Pattern in Omni-Direction Graph Convolution Neural Network architecture, called LKPO-GNN. Our method converts the unordered 3D point cloud into an ordered 1D sequence, to facilitate feeding the raw data into neural networks and simultaneously reducing the computational complexity. LKPO-GNN selects multi-directional k-NNs to form the local topological structure of a centroid, which describes local shapes in the point cloud. Afterwards, GNN is used to combine the local spatial structures and represent the unordered point clouds as a global graph. Experiments on ModelNet40, ShapeNetPart, ScanNet, and S3DIS datasets demonstrate that our proposed method outperforms most existing methods, which verifies the effectiveness and advantage of our work. Additionally, a deep analysis towards illustrating the rationality of our approach, in terms of the learned the topological structure feature, is provided. Source code is available at https://github.com/zwj12377/LKPO-GNN.git. [ABSTRACT FROM AUTHOR]

Facial pose variation is one of the major factors making face recognition (FR) a challenging task. One popular solution is to convert non-frontal faces to frontal ones on which FR is performed. Rotating faces causes facial pixel value changes. Therefore, existing CNN-based methods learn to synthesize frontal faces in color space. However, this learning problem in a color space is highly non-linear, causing the synthetic frontal faces to lose fine facial textures. In this paper, we take the view that the nonfrontal-frontal pixel changes are essentially caused by geometric transformations (rotation, translation, and so on) in space. Therefore, we aim to learn the nonfrontal-frontal facial conversion in the spatial domain rather than the color domain to ease the learning task. To this end, we propose an appearance-flow-based face frontalization convolutional neural network (A3F-CNN). Specifically, A3F-CNN learns to establish the dense correspondence between the non-frontal and frontal faces. Once the correspondence is built, frontal faces are synthesized by explicitly “moving” pixels from the non-frontal one. In this way, the synthetic frontal faces can preserve fine facial textures. To improve the convergence of training, an appearance-flow-guided learning strategy is proposed. In addition, generative adversarial network loss is applied to achieve a more photorealistic face, and a face mirroring method is introduced to handle the self-occlusion problem. Extensive experiments are conducted on face synthesis and pose invariant FR. Results show that our method can synthesize more photorealistic faces than the existing methods in both the controlled and uncontrolled lighting environments. Moreover, we achieve a very competitive FR performance on the Multi-PIE, LFW and IJB-A databases. [ABSTRACT FROM AUTHOR]

Accurate vascular geometry modeling is an essential task in computer assisted vascular surgery and therapy. This paper presents a vessel cross-section based implicit vascular modeling technique, which represents a vascular surface as a set of locally fitted implicit surfaces. In the proposed method, a cross-section based technique is employed to extract from each cross-section of the vascular surface a set of points, which are then fitted with an implicit curve represented as 2D radial basis functions. All these implicitly represented cross-section curves are then being considered as 3D cylindrical objects and combined together using a certain partial shape-preserving spline to build a complete vessel branch; different vessel branches are then blended using a extended smooth maximum function to construct the complete vascular tree. Experimental results show that the proposed method can correctly represent the morphology and topology of vascular structures with high level of smoothness. Both qualitative comparison with other methods and quantitative validations to the proposed method have been performed to verify the accuracy and smoothness of the generated vascular geometric models. [ABSTRACT FROM AUTHOR]

Recovering facial albedo from low quality face images is a challenging task which arises when face recognition is attempted in the wild. Low quality of facial images is usually caused by extrinsic factors such as low resolution and noises, and intrinsic ones such as expressions. Existing research recovers facial albedo by dealing with the extrinsic and intrinsic factors separately. However, it is more natural and potentially more useful to approach albedo recovery by removing the two effects simultaneously. In this paper, we present a novel framework which can recover facial albedo by jointly solving these for both the extrinsic and intrinsic sources of uncertainty. This framework models albedo recovery problem by a joint optimization process which alternatively (1) removes intra-personal variations and (2) performs super resolution. To deal with the intrinsic sources of albedo variability, we use a linear model. To handle extrinsic problems associated with low quality imaging, we use a sparse coding method which is applied to super resolution. The proposed method can also significantly improve the performance of face recognition and clustering in case of very low resolution and in the presence of various facial variations. Extensive experiments and comparisons are conducted on the AR and FERET face databases. Experimental results show the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]

• Considering networks' heterogeneity in embedding methods is effective. • Triangle motifs preserve network heterogeneity and link directionality. • Atomic-level motif embedding overcomes manual intervention. • Higher-order heterogeneous connectivity patterns cope cold-start well. In recent years, there has been a surge of interest in applying machine learning to graphs and networks that already exist in the world around us. The approach has been successfully used for domains as diverse as traffic management, e-commerce recommendation and public opinion monitoring. A critical aspect of representation learning for applied machine learning is feature engineering. Deep learning-based research in representation learning has developed methods for automatically learning a large number of potentially correlated features from original networks. However, most of these methods cannot be applied to heterogeneous networks, which are true expressions of the real-world. This is because they do not adequately capture the structure and semantics of different types of nodes in heterogeneous networks and the links between them. They also struggle to represent higher-order heterogeneous patterns of connection. This paper proposes a generalized motif-based higher-order representation learning method, MBRep, that learns triangle motif embedding in a network, on the basis of which it can obtain the embedding and representation of nodes in a heterogeneous network. Statistically, significant motif structures are extracted from the original heterogeneous network and combined with the heterogeneity of the nodes. A weight-biased random walk is then applied to the motif level higher-order network, using a SkipGram model to embed the motifs. After this, the embedding of the original network nodes is calculated using weighted averages and feature alignment. This can then be used for link prediction. We confirmed the effectiveness of MBRep by comparing its AUC and MRR performance with other state-of-the-art methods on three real-world datasets. Its adaptability was also validated by conducting a cold-start test. [ABSTRACT FROM AUTHOR]

• Replace hard codematrix with soft codematrix. • Use the regression outputs of the base classifiers to replace the traditional classification outputs. • Iteratively softens codematrix elements using global loss tuning. • Conveniently merge the proposed algorithm into other ECOC algorithms to further improve performance. The technique of error-correcting output codes (ECOC) has been proven to be of high discriminative ability in many classification applications. However, most algorithms on the ECOC were designed based on the binary or ternary codes (referred to as the hard codes), which might fail to precisely correct errors in dealing with tough tasks. In this study, a Soft-Recoding strategy based on a self-adaptive algorithm is proposed, which replaces the traditional hard codes with the real-value elements to better fit the output distribution of the base learners. This is achieved by minimizing the ratio of two distances: the distance of the output vector to the ground-truth class, and the average distance of the output vector to the remaining classes. Extensive experiments using five different hard ECOC algorithms and the corresponding softened versions on twenty datasets with diversified numbers of features and classes confirm the effectiveness of our Soft-Recoding strategy in promoting the performance of the original ECOC algorithms. Our source code and additional results are available at: github.com/MLDMXM2017/SA-soft-recoding. [ABSTRACT FROM AUTHOR]

• A new tubular shape modeling technique based on implicit extrusion surfaces (IES) is developed for the accurate geometric modeling of vasculatures. • An implicit geometric editing technique is developed to modify the constructed vascular model with pathology for virtual stenting. • We conduct some blood flow simulations based on the reconstructed and corrected vessel models to illustrate the impact of blood flow dynamics on the diseased vascular wall and on the modified vascular model. High-quality vascular modeling is crucial for blood flow simulations, i.e., computational fluid dynamics (CFD). As without an accurate geometric representation of the smooth vascular surface, it is impossible to make meaningful blood flow simulations. The purpose of this work is to develop high-quality vascular modeling and modification method for blood flow computations. We develop a new technique for the accurate geometric modeling and modification of vasculatures using implicit extrusion surfaces (IES). In the proposed method, the skeleton of the vascular structure is subdivided into short curve segments, each of which is then represented implicitly locally as the intersection of two mutually orthogonal implicit surfaces defined by distance functions. A set of contour points is extracted and fitted with an implicit curve for accurately specifying the vessel cross-section profile, which is then extruded locally along the skeleton to fill the gaps between two vascular tube cross sections. We also present a new implicit geometric editing technique to modify the constructed vascular model with pathology for virtual stenting. Experimental results and validations show that accurate vascular models with highly smooth surfaces can be generated by the proposed method. In addition, we conduct some blood flow simulations to indicate the effectiveness of proposed method for hemodynamic simulations. The proposed technique can achieve precise geometric models of vasculatures with any required degree of smoothness for reliable blood flow simulations. [ABSTRACT FROM AUTHOR]

• We concentrate on the problem of describing the directed flow of information between nodes based on transfer entropy. • We have developed a weighted directed supergraph based on the von Neumann entropy of a directed graph. • Our model can improve the classification performance on fMRI brain connectivity data when the training data are limited. In this paper we develop a novel framework for inferring a generative model of network structure representing the causal relations between data for a set of objects characterized in terms of time series. To do this we make use of transfer entropy as a means of inferring directed information transfer between the time-series data. Transfer entropy allows us to infer directed edges representing the causal relations between pairs of time series, and has thus been used to infer directed graph representations of causal networks for time-series data. We use the expectation maximization algorithm to learn a generative model which captures variations in the causal network over time. We conduct experiments on fMRI brain connectivity data for subjects in different stages of the development of Alzheimer's disease (AD). Here we use the technique to learn class exemplars for different stages in the development of the disease, together with a normal control class, and demonstrate its utility in both graph multi-class and binary classifications. These experiments are showing the effectiveness of our proposed framework when the amounts of training data are relatively small. [ABSTRACT FROM AUTHOR]

Highlights • We proposed a novel ECOC algorithm for multiclass microarray data classification based on the data complexity theory. • Various data complexity measures are deployed to detect the intrinsic characteristics of microarray data sets, so as to produce diverse coding matrices. • A new data complexity measure, named as C1, is designed to evaluate data distribution. It benefits the optimization process of our class partition. • The proposed ECOC algorithm performs more stably in most multiclass microarray data sets compared with other popular ECOC algorithms. Abstract Nowadays, a lot of new classification and clustering techniques have been proposed for microarray data analysis. However, the multiclass microarray data classification is still regarded as a tough task because of the small sample size problem and the class imbalance problem. In this paper, we propose a novel error correcting output code (ECOC) algorithm for the classification of multiclass microarray data based on the data complexity (DC) theory. In this algorithm, an ECOC coding matrix is generated based on a hierarchical partition of the class space with the aim of Minimizing Data Complexity (named as ECOC-MDC). As the partition process can be mapped as a binary tree, a compact ensemble with high discrimination power is produced. The performance of ECOC-MDC is compared with some state-of-art ECOC algorithms on six multiclass microarray data sets, and it is found that the proposed algorithm can obtain better results in most cases. The correlation between DC measures and the dichotomizers' performances is checked, and the observations confirm that high complexity in data usually leads to high error rates of the connected dichotomizers. But the error correcting mechanism in the ECOC framework can effectively improve our algorithm's generalization ability. In short, ECOC-MDC can produce a compact ensemble system with high error correction capability through the application of diverse DC measures. Our Matlab code is available at: github.com/MLDMXM2017/ECOC-MDC. [ABSTRACT FROM AUTHOR]

Abstract Endoscopic ultrasonography (EUS) has shown great advantages in the diagnosis and staging of gastrointestinal malignant tumors. However, EUS based diagnosis is limited by variability in the examiner's subjective interpretation to differentiate between normal and early esophageal carcinoma. In this paper, we propose a novel approach aiming at automatic abnormal region detection from the esophageal EUS images; the contribution is three-fold: first, we present a series of preprocessing strategies developed specifically for the enhancement of EUS images to aid the estimation in the subsequent works. Second, we provide an automatic layer segmentation method based on the multiple surface graph searching approach with incorporation of geometric constraints, which is applied to segment the EUS images into five discernible layers. Third, we introduce the novel feature extraction strategy by utilizing the features from each column in the segmented layers. The SVM classifier is then applied to fulfill the normal and early esophageal carcinoma classification. Subsequently, a clustering method is used to assemble the abnormal columns together so as to detect the abnormal region. Experimental results show that our method has demonstrated its robustness even facing noisy EUS images, and has achieved high accuracy in detecting abnormal region. [ABSTRACT FROM AUTHOR]

The nature of mobile ad hoc networks (MANETs) makes simulation research provide invaluable support for investigating mobile networking protocols, services and applications. Mobility is one of the main factors in simulation of MANETs, due to the fact that it has a strong impact on the design and performance of the networks. Mobility modeling has been an active field for the past decade, mostly focusing on matching a specific mobility or encounter metric with little focus on matching irregular obstacle constraints in realistic scenarios. Consequently, the existing mobility models (MMs) are almost unrealistic. On the other hand, the lack of systemic evaluation framework for MMs makes the mobility characteristics of MANETs be not properly evaluated. In this paper, a realistic mobility model based on Bezier curves (RMBC) is presented. The model operates in an irregular obstacle environment which restricts node movement and wireless transmission. In the RMBC model, a mobile node can calculate smooth pathways between the obstacles using the Bezier curve characterized by control points. Moreover, the flexible movement manners and the realistic application characteristics determined by RMBC are derived and analyzed, respectively. In order to effectively compare the proposed MM with several classical MMs, an integrated and systemic evaluation framework with a multi-dimensional mobility metric space is achieved. The simulation using NS2 tool is conducted. The results show that the proposed MM performs significantly better than the existing MMs in terms of mobility characteristics of MANETs in realistic scenarios. [ABSTRACT FROM AUTHOR]

Due to the high complexity of vascular system network, the geometry reconstruction of vasculatures from raw medical datasets remains a very challenging task. In this paper, we present a novel skeleton-based method for the geometry reconstruction of vascular structures from standard 3D medical datasets. With the proposed techniques, the geometry of vascular structures with high level of smoothness and accuracy can be reconstructed from the raw medical datasets. The experimental results and comparison with other techniques demonstrate that our method can achieve faithful and smooth vascular structures. In addition, quantitative validation has been conducted to evaluate the accuracy and smoothness of the reconstructed vessel geometry based on the proposed method. [ABSTRACT FROM AUTHOR]

In recent years, different computing models have emerged within the area of Unconventional Computation, and more specifically within Natural Computing, getting inspiration from mechanisms present in Nature. In this work, we incorporate concepts in virology and theoretical computer science to propose a novel computational model, called Virus Machine . Inspired by the manner in which viruses transmit from one host to another, a virus machine is a computational paradigm represented as a heterogeneous network that consists of three subnetworks: virus transmission, instruction transfer, and instruction-channel control networks. Virus machines provide non-deterministic sequential devices. As number computing devices, virus machines are proved to be computationally complete, that is, equivalent in power to Turing machines. Nevertheless, when some limitations are imposed with respect to the number of viruses present in the system, then a characterization for semi-linear sets is obtained. [ABSTRACT FROM AUTHOR]