In this paper, we present a new deep learning-based human activity recognition technique. First, we track and extract human body from each frame of the video stream. Next, we abstract human silhouettes and use them to create binary space-time maps (BSTMs) which summarize human activity within a defined time interval. Finally, we use convolutional neural network (CNN) to extract features from BSTMs and classify the activities. To evaluate our approach, we carried out several tests using three public datasets: Weizmann, Keck Gesture and KTH Database. Experimental results show that our technique outperforms conventional state-of-the-art methods in term of recognition accuracy and provides comparable performance against recent deep learning techniques. It’s simple to implement, requires less computing power, and can be used for multi-subject activity recognition.

Characterization of microembolic behavior, as solid or gaseous, guides to an efficient treatment protocol. In this study a new methodology to classify microembolic signals by Deep Convolutional Neural Networks (CNN) is implemented. The experimental system is made up of a flow phantom (ATSLaB) with a cylinder of 6 mm in width. Contrast agents composed of bubbles are employed in this investigational study to imitate the ultrasonic characteristics of gaseous emboli. A Doppler liquid which contains particles, have scatter proprieties analogous to red blood cells, is exploited to mimic the ultrasonic characteristics of the solid emboli. In order to optimize the CNN topology in the training phase, an adaptive learning Root Mean Square (RMSProp) algorithm is used. A classification rate of 99.9% is achieved in this experimental study. These results demonstrate that the CNN optimized model can be adequately exploited for microemboli classification using radio frequency (RF) signals compared to artificial neural networks (ANN) models.

Embolic phenomena, whether air or particulate emboli, can induce immediate damages like heart attack or ischemic stroke. Embolus composition (gaseous or particulate matter) is vital in predicting clinically significant complications. Embolus detection using Doppler methods have shown their limits to differentiate solid and gaseous embolus. Radio-frequency (RF) ultrasound signals backscattered by the emboli contain additional information on the embolus in comparison to the traditionally used Doppler signals. Gaseous bubbles show a nonlinear behavior under specific conditions of the ultrasound excitation wave, this nonlinear behavior is exploited to differentiate solid from gaseous microemboli. In order to verify the usefulness of RF ultrasound signal processing in the detection and classification of microemboli, an in vitro set-up is developed. Sonovue micro bubbles are exploited to mimic the acoustic behavior of gaseous emboli. They are injected at two different concentrations (0.025 and 0.05 µl/ml) in a nonrecirculating flow phantom containing a tube of 0.8 mm in diameter. The tissue mimicking material surrounding the tube is chosen to imitate the acoustic behavior of solid emboli. Both gaseous and solid emboli are imaged using an Anthares ultrasound scanner with a probe emitting at a transmit frequency of 1.82 MHz and at two mechanical indices (MI) 0.2 and 0.6. We propose in this experimental study to exploit discrete wavelet transform and a dimensionality reduction algorithm based on differential evolution technique in the analysis and the characterization of the backscattered RF ultrasound signals from the emboli. Several features are evaluated from the detail coefficients. It should be noted that the features used in this study are the same used in the paper by Aydin et al. These all features are used as inputs to the classification models without using feature selection method. Then we perform feature selection using differential evolution algorithm with support vector machines classifier. The experimental results show clearly that our proposed method achieves better average classification rates compared to the results obtained in a previous study using also the same backscatter RF signals.

In radiotherapy using 18-fluorodeoxyglucose positron emission tomography (¹⁸F-FDG-PET), the accurate delineation of the biological tumour volume (BTV) is a crucial step. In this study, the authors suggest a new approach to segment the BTV in ¹⁸F-FDG-PET images. The technique is based on the k-means clustering algorithm incorporating automatic optimal cluster number estimation, using intrinsic positron emission tomography image information. Clinical dataset of seven patients have a laryngeal tumour with the actual BTV defined by histology serves as a reference, were included in this study for the evaluation of results. Promising results obtained by the proposed approach with a mean error equal to (0.7%) compared with other existing methods in clinical routine, including fuzzy c-means with (35.58%), gradient-based method with (19.14%) and threshold-based methods.

The path-planning problem is commonly formulated to handle the obstacle avoidance constraints. This problem becomes more complicated when further restrictions are added. It often requires efficient algorithms to be solved. In this paper, a new approach is proposed where the path is described by means of Non Uniform Rational B-Splines (NURBS for short) with more additional constraints. An evolutionary technique called Particle Swarm Optimization (PSO) with three options of particles velocity updating offering three alternatives namely the PSO with inertia weight (PSO-W), the constriction factor PSO (PSO-C) and the combination of the two(PSO-WC); are used to optimize the weights of the control points that serve as parameters of the algorithm describing the path. Simulation results show how the mixture of the first two options produces a powerful algorithm, specifically (PSO-WC), in producing a compromise between fast convergence and large number of potential solution. In addition, the whole approach seems to be flexible, powerful and useful for the generation of successful smooth trajectories for robot manipulator which are independent from environment conditions.

In this paper, we present a simple new human activity recognition method using discrete cosine transform (DCT). The scheme uses the DCT coefficients extracted from silhouettes as descriptors (features) and performs frame-by-frame recognition, which make it simple and suitable for real time applications. We carried out several tests using radial basis neural network (RBF) for classification, a comparative study against stat-of-the-art methods shows that our technique is faster, simple and gives higher accuracy performance comparing to discrete transform based techniques and other methods proposed in literature.

Wavelets are known particularly to be an effective tool for extracting discriminative features in the scattered RF signals of both solid and gaseous emboli. However, the selection of an appropriate mother wavelet for the signal being analyzed is an important criterion. This offers the possibility to perform an optimization procedure to obtain the best wavelet. The purpose of the study is to propose a new approach to classify microembolic echoes using a discrete wavelet transform (DWT) based on genetic algorithm optimization and support vector machine (SVM) classifier. The experimental setup consists of a flow phantom (ATSLaB) containing a tube of 6 mm in diameter. In order to mimic the ultrasonic behavior of gaseous emboli, contrast agents consisting of microbubbles are used in our experimental setup. However, to mimic the behavior of the solid emboli we have used the Doppler fluid which contains particles with scatter characteristics comparable to red blood cells. The acquisitions are carried out at 2 MHz and 3.5 MHz transmit frequency. Ultrasound waves are transmitted at different intensities corresponding to mechanical indices (MI) of 0.21 and 0.42 for the transmit frequency of 2 MHz, and 0.31 and 0.62 for the transmit frequency of 3.5 MHz. Two concentrations of the contrast agent (100 μl and 200 μl) are diluted into a 100 ml volume of water. The polyphase representation of the discrete wavelet transform (DWT) is exploited in this study. Such representation allows generating a wavelet filter bank from a set of angular parameters, in order to minimize the fitness function based on genetic algorithm optimization and the SVM classifier. The best accuracy classifications of microemboli obtained in this study are equal to 99.90% for 2MHz and to 99.60% for 3.5MHz. These results illustrate that wavelet optimization approach works well for microemboli classification using RF signals.

Positron emission tomography (PET) plays an important role in early tumour recognition, diagnosis and treatment. Automated and more accurate biological tumour volume (BTV) detection and delineation from PET is challenging. In this paper, we proposed a new method to segment (BTV) in ¹⁸ F-FDG-PET images using an automatic Gaussian mixture model (GMM) based on Akaike information criteria (AIC). The algorithm has been validated on two patients from seven had laryngeal tumours. The volumes estimated were compared with the macroscopic laryngeal specimens in which a 3-D biological tumour volume (BTV) defined by histology served as reference. Experimental results demonstrated that our method was able to segment the (BTV) more accurately than other threshold-based methods.

A common approach to resolve the problem of inverse kinematics of manipulators is based on the Jacobian matrix. However, depending on the complexity of the system to model the elements of the Jacobian matrices may not be calculated. To overcome intrinsic problems related to Jacobian matrix based methods, a new inverse kinematic modeling approach capable to approximate the inverse kinematics of a class of hyper-redundant continuum robots, namely Compact Bionic Handling Assistant (CBHA) is proposed in the present work. The proposed approach makes use of Multilayer Perceptron (MLP) and Radial Basis Function (RBF) Neural Networks as approximation methods. A validation using a rigid 6 DOF industrial manipulator demonstrates the effectiveness and efficiency of the proposed approach.

This paper introduces the active learning approach for wind speed prediction. The main objective of active learning is to opportunely collect training samples in such a way as to minimize the error of the prediction process while minimizing the number of training samples used, and thus to reduce the costs related to the training sample collection. In particular, we propose three different active learning strategies, developed for kernel ridge regression (KRR). The first strategy uses a pool of regressors in order to select the samples with the greater disagreements between the different regressors, while the second one relies on the idea to add samples that are distant from the available training samples, and the last strategy is based on the selection of samples which exhibit a high expected prediction error. A thorough experimental study is presented. It is based on ten different wind speed measurement stations distributed over the vast Algerian territory. Promising results are reported, showing that a smart collection of training samples can be of benefit for wind speed prediction problems.

In this work, we present a detailed experimental assessment of an interesting regression approach based on support vector machines (SVMs), a technique relatively recently introduced in the literature. The experimental framework reports a thorough investigation of the performance of SVMs from different viewpoints, including: (i) the influence of the kernel type in the SVM regression task; (ii) the sensitivity to the number of input variables (spectra dimension); (iii) the sensitivity to the available number of training samples; and (iv) the overall stability. The obtained results are compared with those yielded by the radial basis function (RBF) and the multilayer perceptron (MLP) neural networks as well as the traditional multiple linear regression (MLR) method on two different spectrophotometric datasets.

The effectiveness of a regression method strongly depends on the characteristics of the considered regression problem. As a consequence, this makes it difficult to choose a priori the most appropriate algorithm for a given dataset. This issue is faced in this work through a novel regression approach based on the fusion of an ensemble of different regressors. In order to implement the proposed robust multiple system (RMS), four different fusion strategies are explored. In this context, we propose a novel fusion strategy named selection‐based strategy (SBS) that provides as output the estimate obtained by the regression algorithm (included in the ensemble) characterized by the highest expected accuracy in the region of the feature space associated with the considered model. The SBS is based not on a direct combination of the estimates yielded by all the regressors but on a selection mechanism that identifies the expected best available estimate. For such purpose, it exploits the accuracies of the regressors included in the ensemble in different portions of the input feature space. The experimental assessment of the RMS was carried out on three different datasets: a wine, an orange juice, and an apple datasets. The obtained experimental results suggest that, in general, the fusion of an ensemble of different regression algorithms leads to a regression process that is more robust and sometimes also more accurate than traditional regression methods. In particular, the proposed SBS method represents an effective solution to carry out the fusion process.

In this work, we introduce an active learning approach for the estimation of chemical concentrations from spectroscopic data. Its main objective is to opportunely collect training samples in such a way as to minimize the error of the regression process while minimizing the number of training samples used, and thus to reduce the costs related to training sample collection. In particular, we propose two different active learning strategies developed for regression approaches based on partial least squares regression, ridge regression, kernel ridge regression, and support vector regression. The first strategy uses a pool of regressors in order to select the samples with the greatest disagreements among the different regressors of the pool, while the second one is based on adding samples that are distant from the current training samples in the feature space. For support vector regression, a specific strategy based on the selection of the samples distant from the support vectors is proposed. Experimental results on three different real data sets are reported and discussed.

This paper considers the design of a lossy image compression algorithm dedicated to color still images. After a preprocessing step (mean removing and RGB to YCbCr transformation), the DCT transform is applied and followed by an iterative phase (using the bisection method) including the thresholding, the quantization, dequantization, the inverse DCT, YCbCr to RGB transform and the mean recovering. This is done in order to guarantee that a desired quality (fixed in advance using the well known PSNR metric) is checked. For the aim to obtain the best possible compression ratio CR, the next step is the application of a proposed adaptive scanning providing, for each (n, n) DCT block a corresponding (n×n) vector containing the maximum possible run of zeros at its end. The last step is the application of a modified systematic lossless encoder. The efficiency of the proposed scheme is demonstrated by results, especially, when faced to the method presented in the recently published paper based on the block truncation coding using pattern fitting principle.

Emboli classification is of high clinical importance for selecting appropriate treatment for patients. Several ultrasonic (US) methods using Doppler processing have been used for emboli detection and classification as solid or gaseous matter. We suggest in this experimental study exploiting the Radio-Frequency (RF) signal backscattered by the emboli since they contain additional information on the embolus than the Doppler signal. The aim of the study is the analysis of RF signals using Multilayer Perceptron (MLP) and Radial-Basis Function Network (RBFN) in order to classify emboli.

Anthares scanner with RF access was used with a transmit frequency of 1.82 MHz at two mechanical indices (MI) 0.2 and 0.6. The mechanical index is given as the peak negative pressure (in MPa) divided by the square root of the frequency (in MHz). A Doppler flow phantom was used containing a 0.8 mm diameter vessel surrounded by a tissue mimicking material. To imitate gas emboli US behaviour, Sonovue microbubbles were injected at two different doses (10μl and 5μl) in a nonrecirculating at a constant flow. The surrounding tissue was assumed to behave as a solid emboli. In order to mimic real clinical pathological situations, Sonovue concentration was chosen such that the fundamental scattering from the tissue and from the contrast were identical. The amplitudes and bandwidths of the fundamental and the 2nd harmonic components were selected as input parameters to the MLP and RBFN models. Moreover the frequency bandwidths of the fundamental and the 2nd harmonic echoes were approximated by Gaussian functions and the coefficients were used as a third input parameter to the neural network models. The results show that the Gaussian coefficients provide the highest rate of classification in comparison to the amplitudes and the bandwidths of the fundamental and the 2nd harmonic components. The classification rates reached 89.28% and 92.85% with MLP and RBFN models respectively.

This short communication demonstrates the opportunity to classify emboli based on a RF signals and neural network analysis.

A new approach based on multiple architecture system (MAS) for the prediction of wind speed is proposed. The motivation behind the proposed approach is to combine the complementary predictive powers of multiple models in order to improve the performance of the prediction process. The proposed MAS can be implemented by associating the predictions obtained from the different regression algorithms (MLR, MLP, RBF and SVM) making up the ensemble by three fusion strategies (simple, weighted and non-linear). The efficiency of the proposed approach has been assessed on a real data set recorded from seven locations in Algeria during a period of 10 years. The experimental results point out that the proposed MAS approach is capable of improving the precision of the wind speed prediction compared to the traditional prediction methods.

In this paper, we propose a two-stage regression approach, which is based on the residual correction concept. Its underlying idea is to correct any given regressor by analyzing and modeling its residual errors in the input space. We report and discuss results of experiments conducted on three different datasets in infrared spectroscopy and designed in such a way to test the proposed approach by: 1) varying the kind of adopted regression method used to approximate the chemical parameter of interest. Partial least squares regression (PLSR), support vector machines (SVM) and radial basis function neural network (RBF) methods are considered; 2) adopting or not a feature selection strategy to reduce the dimension of the space where to perform the regression task. A comparative study with another approach which exploits differently estimation errors, namely adaptive boosting for regression (AdaBoost.R), is also included. The obtained results point out that the residual-based correction approach (RBC) can improve the accuracy of the estimation process. Not all the improvements are statistically significant but, at the same time, no case of accuracy decrease has been observed.

Over the last few decades medical ultrasound has become an established diagnostic and therapeutic tool since it produces detailed and high resolution images of tissues in human body. Harmonic imaging is among the recent developments which has brought further improvements to the image quality. Harmonic imaging for tissue or with contrast agent induced a rapid evolution of this modality to diverse clinical uses, among which myocardial perfusion determination seems to be the most important application. This brought the need to understand the physical processes involved in the propagation of finite amplitude sound beams, and the issues for redesigning and optimizing the transducers with higher performances for both tissue imaging and contrast imaging. Concerning tissue harmonic imaging, the advantage of the harmonic beam generated at two times the transmit frequency are translated by reduced reverberations , greater depth of penetration at higher frequencies and improved resolution. In order to characterize the harmonic beam, a time domain solution of the parabolic nonlinear wave equation is used. This equation is traditionally applied in a propagation direction along the central transducer axis, and has been shown to model the pulse propagation satisfactorily. In this work, the characteristics and performances of the second harmonic acoustic beam from a focused piston aperture are described and the physical principles behind tissue harmonic imaging are computed. The field properties are then discussed regarding image quality. Special attention is given to the transmitted and received bandwidths variation and the reception of the pure echo signal.