The growth of prebreakdown cavities in silicone fluids and thefrequency of the accompanying discharge pulses

Measurements have been made of prebreakdown cavities in silicone fluids, and of the current pulses that accompany cavity growth. These experiments were carried out in silicone fluids of 0.65, 10, 100 and 1000 cS viscosity. Cavity growth, driven by the electrostatic field, is limited at low viscosities by inertia, and at high viscosities by viscous drag. The electrostatic force on the cavity wall is related to the local field and to the space charge density in the liquid adjacent to the cavity. We are concerned with the relationship between the electrostatic force and the cavity growth, and with the discharges that accompany cavity growth. Discharges occur in well-defined pulse trains: the first pulse in a train generates the cavity, and subsequent pulses are due to discharges within the cavity. Knowing the scaling laws for cavity growth we can use the time between the first and second pulses to estimate the cavity size when the first cavity discharge occurs; this gives a cavity diameter of ~5 to 7 μm. The next pulse cannot occur until the charge from the previous discharge has dispersed. We find that the time between pulses Δt is strongly viscosity dependent; at high viscosities the average time between pulses at is proportional to fluid-viscosity, but in the low viscosity limit the dependence approaches η^1/3. To explain this viscosity dependence we consider three mechanisms: (1) a decrease in charge density due to increase in cavity size; (2) ion detrapping from the cavity wall and drift in the applied field; and (3) diffusion of an impurity species to the cavity surface, charge exchange to create a mobile ion, and its subsequent drift in the field. Our experimental results are consistent with the cavity expansion model, but there is evidence of diffusion effects in low viscosity liquids, and with ion-drift at high viscosities     

Adaptive transfer learning in deep neural networks: Wind power prediction using knowledge transfer from region to region and between different task domains

Transfer learning (TL) in deep neural networks is gaining importance because, in most of the applications, the labeling of data is costly and time consuming. Additionally, TL also provides an effective weight initialization strategy for deep neural networks. This paper introduces the idea of adaptive TL in deep neural networks (ATL‐DNN) for wind power prediction. Specifically, we show in case of wind power prediction that adaptive TL of the deep neural networks system can be adaptively modified as regards training on a different wind farm is concerned. The proposed ATL‐DNN technique is tested for short‐term wind power prediction, where continuously arriving information has to be exploited. Adaptive TL not only helps in providing good weight initialization, but also in utilizing the incoming data for effective learning. Additionally, the proposed ATL‐DNN technique is shown to transfer knowledge between different task domains (wind power to wind speed prediction) and from one region to another region. The simulation results show that the proposed ATL‐DNN technique achieves average values of 0.0637, 0.0986, and 0.0984 for the mean absolute error, root mean squared error, and standard deviation error, respectively.     

An effective back‐off selection technique for reliable beacon reception in VANETs

In safety‐critical scenarios, reliable reception of beacons transmitted by a subject vehicle is critical to avoid vehicle collision. According to the employed contention window sizes in IEEE 802.11p, beacons are transmitted with a small contention window size. As a result, multiple vehicles contend for the shared channel access by selecting the same back‐off slot. This is a perfect recipe for synchronous collisions wherein reliable beacon delivery cannot be guaranteed for any vehicle. We consider the problem of selecting the back‐off slots from the current contention window to provide reliable delivery of beacons transmitted by a subject vehicle to its neighbors. Given a safety scenario, we propose a Pseudo‐Random Number Generator (PRNG)‐inspired back‐off selection (PBS) technique. The proposed technique works on the hypothesis that synchronous collisions of beacons transmitted by a subject vehicle can be reduced if all its neighbors select different back‐off slots (ie, not the back‐off slot selected by the subject vehicle). The discrete‐event simulations demonstrate that PBS can increase the overall message reception from a subject vehicle, in comparison with the uniform random probability back‐off selection in IEEE 802.11p.     

GWO‐C: Grey wolf optimizer‐based clustering scheme for WSNs

A wireless sensor network (WSN) is a prominent technology that could assist in the fourth industrial revolution. Sensor nodes present in the WSNs are functioned by a battery. It is impossible to recharge or replace the battery, hence energy is the most important resource of WSNs. Many techniques have been devised and used over the years to conserve this scarce resource of WSNs. Clustering has turned out to be one of the most efficient methods for this purpose. This paper intends to propose an efficient technique for election of cluster heads in WSNs to increase the network lifespan. For the achievement of this task, grey wolf optimizer (GWO) has been employed. In this paper, the general GWO has been modified to cater to the specific purpose of cluster head selection in WSNs. The objective function for the proposed formulation considers average intra‐cluster distance, sink distance, residual energy, and CH balancing factor. The simulations are carried out in diverse conditions. On comparison of the proposed protocol, ie, GWO‐C protocol with some well‐known clustering protocols, the obtained results prove that the proposed protocol outperforms with respect to the consumption of the energy, throughput, and the lifespan of the network. The proposed protocol forms energy‐efficient and scalable clusters.     

Color image segmentation by combining the convex active contour and the Chan Vese model

Pattern Analysis and Applications - In this paper, we present a robust and computationally efficient image segmentation technique based on a hybrid convex active contour and the Chan–Vese...     

A Novel Auto-Annotation Technique for Aspect Level Sentiment Analysis

In machine learning, sentiment analysis is a technique to find and analyze the sentiments hidden in the text. For sentiment analysis, annotated data is a basic requirement. Generally, this data is manually annotated. Manual annotation is time consuming, costly and laborious process. To overcome these resource constraints this research has proposed a fully automated annotation technique for aspect level sentiment analysis. Dataset is created from the reviews of ten most popular songs on YouTube. Reviews of five aspects—voice, video, music, lyrics and song, are extracted. An N-Gram based technique is proposed. Complete dataset consists of 369436 reviews that took 173.53 s to annotate using the proposed technique while this dataset might have taken approximately 2.07 million seconds (575 h) if it was annotated manually. For the validation of the proposed technique, a sub-dataset—Voice, is annotated manually as well as with the proposed technique. Cohen's Kappa statistics is used to evaluate the degree of agreement between the two annotations. The high Kappa value (i.e., 0.9571%) shows the high level of agreement between the two. This validates that the quality of annotation of the proposed technique is as good as manual annotation even with far less computational cost. This research also contributes in consolidating the guidelines for the manual annotation process.     

Self-embedding framework for tamper detection and restoration of color images

Multimedia Tools and Applications - The present era is paving huge expansion to the transmission of digital data in fields like health, military intelligence, scientific research, and publication...     

Underwater object detection: architectures and algorithms – a comprehensive review

Underwater object detection is an essential step in image processing and it plays a vital role in several applications such as the repair and maintenance of sub-aquatic structures and marine sciences. Many computer vision-based solutions have been proposed but an optimal solution for underwater object detection and species classification does not exist. This is mainly because of the challenges presented by the underwater environment which mainly include light scattering and light absorption. The advent of deep learning has enabled researchers to solve various problems like protection of the subaquatic ecological environment, emergency rescue, reducing chances of underwater disaster and its prevention, underwater target detection, spooring, and recognition. However, the advantages and shortcomings of these deep learning algorithms are still unclear. Thus, to give a clearer view of the underwater object detection algorithms and their pros and cons, we proffer a state-of-the-art review of different computer vision-based approaches that have been developed as yet. Besides, a comparison of various state-of-the-art schemes is made based on various objective indices and future research directions in the field of underwater object detection have also been proffered.

     

A novel offline handwritten text recognition technique to convert ruled-line text into digital text through deep neural networks

Multimedia Tools and Applications - Offline Handwritten Text Recognition (HTR) has been an active area of research due to its wide range of applications and challenges. Recently, many offline HTR...     

Sustainable Green Fog Computing for Smart Agriculture

Wireless Personal Communications - Medical Body Area Networks or MBANs are gaining popularity in healthcare circles because of the convenience they provide to patients and caregivers and assist in...