CA : A Cancer Journal for Clinicians

Super-resolution, the process of obtaining one or more high-resolution images from one or more low-resolution observations, has been a very attractive research topic over the last two decades. It has found practical applications in many real-world problems in different fields, from satellite and aerial imaging to medical image processing, to facial image analysis, text image analysis, sign and number plates reading, and biometrics recognition, to name a few. This has resulted in many research papers, each developing a new super-resolution algorithm for a specific purpose. The current comprehensive survey provides an overview of most of these published works by grouping them in a broad taxonomy. For each of the groups in the taxonomy, the basic concepts of the algorithms are first explained and then the paths through which each of these groups have evolved are given in detail, by mentioning the contributions of different authors to the basic concepts of each group. Furthermore, common issues in super-resolution algorithms, such as imaging models and registration algorithms, optimization of the cost functions employed, dealing with color information, improvement factors, assessment of super-resolution algorithms, and the most commonly employed databases are discussed.

/pdf/super-resolution-a-comprehensive-survey-xjl42yt9nt.pdf

Super-resolution: a comprehensive survey

For any positive integer $k$, there exist neural networks with $\Theta(k^3)$ layers, $\Theta(1)$ nodes per layer, and $\Theta(1)$ distinct parameters which can not be approximated by networks with $\mathcal{O}(k)$ layers unless they are exponentially large --- they must possess $\Omega(2^k)$ nodes. This result is proved here for a class of nodes termed "semi-algebraic gates" which includes the common choices of ReLU, maximum, indicator, and piecewise polynomial functions, therefore establishing benefits of depth against not just standard networks with ReLU gates, but also convolutional networks with ReLU and maximization gates, sum-product networks, and boosted decision trees (in this last case with a stronger separation: $\Omega(2^{k^3})$ total tree nodes are required).

Benefits of depth in neural networks

A novel multiple-input multiple-output (MIMO) transmission scheme, called space-time block coded spatial modulation (STBC-SM), is proposed. It combines spatial modulation (SM) and space-time block coding (STBC) to take advantage of the benefits of both while avoiding their drawbacks. In the STBCSM scheme, the transmitted information symbols are expanded not only to the space and time domains but also to the spatial (antenna) domain which corresponds to the on/off status of the transmit antennas available at the space domain, and therefore both core STBC and antenna indices carry information. A general technique is presented for the design of the STBC-SM scheme for any number of transmits antennas. Besides the high spectral efficiency advantage provided by the antenna domain, the proposed scheme is also optimized by deriving its diversity and coding gains to exploit the diversity advantage of STBC. A low-complexity maximum likelihood (ML) decoder is given for the new scheme which profits from the orthogonality of the core STBC. The performance advantages of the STBC-SM over simple SM and over V-BLAST are shown by simulation results for various spectral efficiencies and are supported by the derivation of a closed form expression for the union bound on the bit error probability.

Space-Time Block Coded Spatial Modulation

Machine learning techniques for breast cancer computer aided diagnosis using different image modalities: A systematic review.

Breast tumor classification in ultrasound images using texture analysis and super-resolution methods

A two-dimensional image segmentation method based on genetic algorithm and entropy

This paper proposed a novel deep learning-based relay selection scheme in millimeter wave (mmWave) Device-to-Device (D2D) communication underlying the fifth generation (5G) cellular networks. Relay selection seems to be a promising solution to extend the coverage and solve the blocking problem of mmWave direct communication. In the case of direct path blocking, the base station (BS)/user equipment (UE) has several candidate devices to be selected as a relay. Despite that, conventional scheme when the direct path is blocked, the direct communication link is handover from mmWave to a lower frequency band using a fast session transfer (FST) technique. Such a blocking problem can be solved multi-hop communications by relaying data based on select another device as a relay. Motivated by the importance of selecting the optimal relay which increases reliable connectivity in mmWave communication and expansion of coverage. The proposed deep learning model is developed to overcome the challenges of selected the optimal relay based with low complexity and high efficiency. The proposed scheme considers a deep learning model learns how to predict the best relay for relaying the data in high-reliability communication. The in-depth learning approach is recommended due to its capability in constructing an intelligent model that can take successful decisions and make precise predictions. Simulation results show that the proposed relay selection algorithm outperforms the conventional relay selection in D2D technique in the spectral efficiency and the energy efficiency

Deep Learning-Based Relay Selection In D2D Millimeter Wave Communications

Choosing a proper Region of Interest (ROI) for Remote Photoplethysmography (rPPG) is essential and a challenging first step, and it has a direct effect on the accuracy and reliability of the overall heart rate (HR) algorithm. Non-skin areas have no contribution to the HR information; however, few works have tackled the issue of non-skin pixels included in the ROI. First, this paper considers improving the quality of the rPPG signal by filtering out non-skin pixels included within the ROI. The feasibility of employing skin segmentation for ROI definition is demonstrated. Then, this technique is compared with our previous real-time rPPG-based method. Moreover, we explore the effect of extracting the HR from three ROIs using signal fusion. Second, we give a comprehensive account of the examined methods in our algorithm for face detection, face tracking, skin detection, and blind signal separation. Finally, we compare our rPPG measurements with ground truth values obtained from a commercial pulse oximeter. Based on the simulation results, the proposed algorrithm significantly improves the quality of the rPPG technique.

/pdf/optimizing-remote-photoplethysmography-using-adaptive-skin-dv7cby0lzm.pdf

Optimizing Remote Photoplethysmography Using Adaptive Skin Segmentation for Real-Time Heart Rate Monitoring

Malignant melanoma is considered as one of the most dangerous type of skin cancers as it increases the mortality rate. Computer-aided diagnosis (CAD) systems can help to detect melanoma early. In this paper, we propose a new skin melanoma CAD system using texture analysis methods. The proposed CAD system consists of four steps: hair removal, filtering, feature extraction and classification. In the feature extraction step, we evaluate the performance of five widely used texture analysis methods: grey level co-occurrence matrix, Gabor filters, a histogram of oriented gradients, local binary pattern and local directional number. Our CAD system classifies between non-melanoma skin lesions (represented as common nevi or dysplastic nevi) and melanoma. The experimental results show that extracting HOG features after hair removal yields the best classification results. HOG gives an AUC of 0.9783 with melanoma/common nevi classification and an AUC of 0.9439 with melanoma/dysplastic nevi classification.

Osama A. Omer

Papers

Breast tumor classification in ultrasound images using texture analysis and super-resolution methods

A two-dimensional image segmentation method based on genetic algorithm and entropy

Deep Learning-Based Relay Selection In D2D Millimeter Wave Communications

Optimizing Remote Photoplethysmography Using Adaptive Skin Segmentation for Real-Time Heart Rate Monitoring

Computer aided diagnosis system for skin lesions detection using texture analysis methods