Modern machine-learning applications require GPUs, and modern platforms can leverage numerous GPUs on one or more machines to increase performance. Contemporary deep-learning models are too huge for CPU or GPU training. Training these models with many GPUs without performance degradation is necessary to train them rapidly and maximize GPU consumption. Thus, training deep convolutional neural networks (DCNN) with multiple GPUs has become necessary for improving training. Therefore, we presented a parallel design and development of an efficient model for enhancing face mask CNN performance and improving resource efficiency. This DCNN model is a parallel training system over multiple GPUs, a multi-core CPU, and a multi-process GPU platform with large batch size and learning rate involvement to optimize resource use across storage, configuration and scaling using large datasets. the proposed model contains two parts, the first one is used for specifying and extracting the faces using the Haar Cascade classifier, and the second one considers the core part that extracts features from facial images for classification. As a result, the average speed of a multi-GPU is about 2.7 times faster than the GPU and about 3.2 times faster than the CPU. Also, according to our evaluation results, the training time obtained using multiple GPUs and multiple processes is much smaller than that obtained using a single GPU single process. Parallel training on multiple GPUs improves GPU resource utilization and training throughput. This model reflects significant accuracy compared to the other commonly used methods from relevant articles by achieving an Accuracy score of 99.5%.
Suicidal ideation is one of the most severe mental health issues faced by people all over the world. There are various risk factors involved that can lead to suicide. The most common & critical risk factors among them are depression, anxiety, social isolation and hopelessness. Early detection of these risk factors can help in preventing or reducing the number of suicides. Online social networking platforms like Twitter, Redditt and Facebook are becoming a new way for the people to express themselves freely without worrying about social stigma. This paper presents a methodology and experimentation using social media as a tool to analyse the suicidal ideation in a better way, thus helping in preventing the chances of being the victim o
... Show MoreAudio classification is the process to classify different audio types according to contents. It is implemented in a large variety of real world problems, all classification applications allowed the target subjects to be viewed as a specific type of audio and hence, there is a variety in the audio types and every type has to be treatedcarefully according to its significant properties.Feature extraction is an important process for audio classification. This workintroduces several sets of features according to the type, two types of audio (datasets) were studied. Two different features sets are proposed: (i) firstorder gradient feature vector, and (ii) Local roughness feature vector, the experimentsshowed that the results are competitive to
... Show MoreWatermarking operation can be defined as a process of embedding special wanted and reversible information in important secure files to protect the ownership or information of the wanted cover file based on the proposed singular value decomposition (SVD) watermark. The proposed method for digital watermark has very huge domain for constructing final number and this mean protecting watermark from conflict. The cover file is the important image need to be protected. A hidden watermark is a unique number extracted from the cover file by performing proposed related and successive operations, starting by dividing the original image into four various parts with unequal size. Each part of these four treated as a separate matrix and applying SVD
... Show MoreThe rapid development of telemedicine services and the requirements for exchanging medical information between physicians, consultants, and health institutions have made the protection of patients’ information an important priority for any future e-health system. The protection of medical information, including the cover (i.e. medical image), has a specificity that slightly differs from the requirements for protecting other information. It is necessary to preserve the cover greatly due to its importance on the reception side as medical staff use this information to provide a diagnosis to save a patient's life. If the cover is tampered with, this leads to failure in achieving the goal of telemedicine. Therefore, this work provides an in
... Show MoreLuminescent solar concentrator (LSC) are used to enhance photoresponsivity of solar cell. The Quantumdots luminescent solar concentrator (QDLSC) consists of CdSe/CdS core/shell nanoparticles embedded in polyacrylamide polymer matrix positioned on the top surface of the silicon solar cell. This procedure improves the conversion efficiency of the bare silicon solar cell. The conversion efficiency of the solar cell has increased from 7.3% to 10.3%. this improvement is referred to the widening of the response spectral region window of the a- Si. Solar cell.
Computations of the relative permeability curves were made through their representation by two functions for wetting and nonwetting phases. Each function contains one parameter that controls the shape of the relative permeability curves. The values of these parameters are chosen to minimize an objective function, that is represented as a weighted sum of the squared differences between experimentally measured data and the corresponding data calculated by a mathematical model simulating the experiment. These data comprise the pressure drop across core samples and the recovery response of the displacing phase. Two mathematical models are constructed in this study to simulate incompressible, one-dimensional, two-phase flow. The first model d
... Show MoreElastic magnetic M1 electron scattering form factor has been calculated for the ground state J,T=1/2-,1/2 of 13C. The single-particle model is used with harmonic oscillator wave function. The core-polarization effects are calculated in the first-order perturbation theory including excitations up to 5ħω, using the modified surface delta interaction (MSDI) as a residual interaction. No parameters are introduced in this work. The data are reasonably explained up to q~2.5fm-1 .
Background: Avascular necrosis (AVN) is defined as cellular death of bone components due to interruption of the blood supply; the bone structures then collapse, resulting in bone destruction, pain, and loss of joint function. AVN is associated with numerous conditions and usually involves the epiphysis of long bones, such as the femoral head. In clinical practice, AVN is most commonly encountered in the hip. Early diagnosis and appropriate intervention can delay the need for joint replacement. However, most patients present late in the disease course. Without treatment, the process is almost always progressive, leading to joint destruction within 5 years.Treatment of a vascular necrosis depends mainly on early diagnosis which mainly base
... Show More