Identification of Anomaly Scenes in Videos Using Graph Neural Networks

Generally, conventional methods for anomaly detection rely on clustering, proximity, or classification. With the massive growth in surveillance videos, outliers or anomalies find ingenious ways to obscure themselves in the network and make conventional techniques inefficient. This research explores the structure of Graph neural networks (GNNs) that generalize deep learning frameworks to graph-structured data. Every node in the graph structure is labeled and anomalies, represented by unlabeled nodes, are predicted by performing random walks on the node-based graph structures. Due to their strong learning abilities, GNNs gained popularity in various domains such as natural language processing, social network analytics and healthcare. Anomaly detection is a challenging task in computer vision but the proposed algorithm using GNNs efficiently performs the identification of anomalies. The Graph-based deep learning networks are designed to predict unknown objects and outliers. In our case, they detect unusual objects in the form of malicious nodes. The edges between nodes represent a relationship of nodes among each other. In case of anomaly, such as the bike rider in Pedestrians data, the rider node has a negative value for the edge and it is identified as an anomaly. The encoding and decoding layers are crucial for determining how statistical measurements affect anomaly identification and for correcting the graph path to the best possible outcome. Results show that the proposed framework is a step ahead of the traditional approaches in detecting unusual activities, which shows a huge potential in automatically monitoring surveillance videos. Performing autonomous monitoring of CCTV, crime control and damage or destruction by a group of people or crowd can be identified and alarms may be triggered in unusual activities in streets or public places. The suggested GNN model improves accuracy by 4% for the Pedestrian 2 dataset and 12% for the Pedestrian 1 dataset compared to a few state-of-the-art techniques.     

Nickel-titanium shape memory alloys made by selective laser melting:a review on process optimisation

Selective laser melting (SLM) is a mainstream powder-bed fusion additive manufacturing (AM) process that creates a three-dimensional (3D) object using a high power laser to fuse fine particles of various metallic powders such as copper, tool steel, cobalt chrome, titanium, tungsten, aluminium and stainless steel. Over the past decade, SLM has received significant attention due to its capability in producing dense parts with superior mechanical properties. As a premier shape memory alloy, the nickel-titanium (NiTi) shape memory alloy is attractive for a variety of biomedical applications due to its superior mechanical properties, superelasticity, corrosion resistance and biocompatibility. This paper presents a comprehensive review of the recent progress in NiTi alloys produced by the SLM process, with a particular focus on the relationship between processing parameters, resultant microstructures and properties. Current research gaps, challenges and suggestions for future research are also addressed.The full text can be downloaded at https://link.springer.com/article/10.1007/s40436-021-00376-9     

Macroscale assessment of low-velocity impact on hybrid composite laminates

Composites are usually brittle materials and have low impact properties. Structural dimensions, stacking sequence, ply materials, ply thicknesses and ply angles are standard variables that influence composite‘s performance against impact loads. Stacking sequence in hybrid laminates affects the failure and impact resistance. Failure mechanisms at the low-velocity impact of a rigid object in hybrid laminates are complex, and the subsurface damage in a composite laminate cannot be detected directly. However, various simulation platforms make it easy to see the impact damage between the plies of laminate. This paper numerically investigated the effect of stack sequence and hybridization of two fiber types against low-velocity impact. The current study adopted four-layer composite laminates of carbon and glass fiber layers with a stacking plan [C/C/C/C], [C/G/C/G] and [G/C/G/C], having lay-up angles as [0°/45°/−45°/90°]. Keeping the impactor mass and the incident velocity constant, the laminates were subjected to low-velocity impact. The damage contours for a failure mode were recorded and compared at the ply level. The numerical study resulted in impact imitations showing comparisons of the damage contours using Hashin failure criteria. Hybrid laminates display better performance in absorbing impact energies; however, hybrid laminates experienced more subsurface damage due to more impact energy absorption.     

A Bio-Inspired Routing Optimization in UAV-enabled Internet of Everything

Internet of Everything (IoE) indicates a fantastic vision of the future, where everything is connected to the internet, providing intelligent services and facilitating decision making. IoE is the collection of static and moving objects able to coordinate and communicate with each other. The moving objects may consist of ground segments and flying segments. The speed of flying segment e.g., Unmanned Ariel Vehicles (UAVs) may high as compared to ground segment objects. The topology changes occur very frequently due to high speed nature of objects in UAV-enabled IoE (Ue-IoE). The routing maintenance overhead may increase when scaling the Ue-IoE (number of objects increases). A single change in topology can force all the objects of the Ue-IoE to update their routing tables. Similarly, the frequent updating in routing table entries will result more energy dissipation and the lifetime of the Ue-IoE may decrease. The objects consume more energy on routing computations. To prevent the frequent updation of routing tables associated with each object, the computation of routes from source to destination may be limited to optimum number of objects in the Ue-IoE. In this article, we propose a routing scheme in which the responsibility of route computation (from neighbor objects to destination) is assigned to some IoE-objects in the Ue-IoE. The route computation objects (RCO) are selected on the basis of certain parameters like remaining energy and mobility. The RCO send the routing information of destination objects to their neighbors once they want to communicate with other objects. The proposed protocol is simulated and the results show that it outperform state-of-the-art protocols in terms of average energy consumption, messages overhead, throughput, delay etc.     

Predicting scientific impact based on <Emphasis Type="Italic">h</Emphasis>-index

Predicting the future impact of a scientist/researcher is a critical task. The objective of this work is to evaluate different h-index prediction models for the field of Computer Science. Different combinations of parameters have been identified to build the model and applied on a large data set taken from Arnetminer comprised of almost 1.8 million authors and 2.1 million publications’ record of Computer Science. Machine learning prediction technique, regression, is used to find the best set of parameters suitable for h-index prediction for the scientists from all career ages, without enforcing any constraint on their current h-index values with R ² as a metric to measure the accuracy. Further, these parameters are evaluated for different career ages and different thresholds for h-index values. Prediction results for 1 year are really good, having R ² 0.93 but for 5 years R ² declines to 0.82 on average. Hence inferred that prediction of h-index is difficult for longer periods. Predictions for the researchers having 1 year experience are not precise, having R ² 0.60 for 1 year and 0.33 for 5 years. Considering scientists of different career ages, average R ² values for researchers having 20–36 years of experience were 0.99. For the researches having different h-index values, researchers having low h-index were difficult to predict. Parameters set comprising of current h-index, average citations per paper, number of coauthors, years since publishing first article, number of publications, number of impact factor publications, and number of publications in distinct journals performed better than all other combinations.     

Plasmonic photothermal therapy of colon cancer cells utilising gold nanoshells: an in vitro study

In this study, gold nanoshell (GNS) were synthesised utilising the Halas method. The obtained nanoparticles (NPs) were characterised by Fourier‐transform infrared (FTIR) spectroscopy, scanning electron microscopy (SEM), transmission electron microscopy (TEM), UV–Vis spectroscopy and dynamic light scattering. FTIR spectra demonstrated the successful functionalisation of silica NP with 3‐aminopropyl trimethoxysilane. SEM and TEM images showed the morphology and diameter of the synthesised silica NPs (137 ± 26 nm) and GNS. UV–Vis spectrum illustrated the maximum absorbance of the resultant GNS and their average hydrodynamic diameter was 159 nm. For in vitro study, HCT‐116 cells were exposed to gold nanoshells and intense pulsed light in different experiment groups. The results showed that exposing the cells to nanoshells and 30 s irradiation would efficiently decrease the viability percentage of the cells to about 30% compared with the control. A continued exposure of 4 min decreased the viability of the cancer cells to 20%. The results demonstrated that photothermal therapy would be promising in treatment of colon cancer cells utilising gold nanoshells.Inspec keywords: gold, silicon compounds, nanomedicine, plasmonics, radiation therapy, bio‐optics, cancer, cellular biophysics, nanoparticles, Fourier transform spectra, infrared spectra, scanning electron microscopy, transmission electron microscopy, ultraviolet spectra, visible spectraOther keywords: plasmonic photothermal therapy, colon cancer cells, gold‐silica nanoshells, GNS, Halas method, Fourier transform infrared spectroscopy, FTIR, scanning electron microscopy, SEM, transmission electron microscopy, TEM, UV‐vis spectroscopy, dynamic light scattering, FTIR spectra, 3‐aminopropyl trimethoxysilane, morphology, in vitro study, HCT‐116 cells, cell viability, nanoparticles, time 30 s, time 4 min, Au     

Automatic Correction of Saturated Regions in Photographs using Cross‐Channel Correlation

Incorrectly setting the camera's exposure can have a significant negative effect on a photograph. Over‐exposing photographs causes pixels to exhibit unpleasant artifacts due to saturation of the sensor. Saturation removal typically involves user intervention to adjust the color values, which is tedious and time‐consuming. This paper discusses how saturation can be automatically removed without compromising the essential details of the image. Our method is based on a smoothness prior: neighboring pixels have similar channel ratios and color values. We demonstrate that high quality saturation‐free photos can be obtained from a simple but effective approach.     

Scalar field visualization via extraction of symmetric structures

Identifying symmetry in scalar fields is a recent area of research in scientific visualization and computer graphics communities. Symmetry detection techniques based on abstract representations of the scalar field use only limited geometric information in their analysis. Hence they may not be suited for applications that study the geometric properties of the regions in the domain. On the other hand, methods that accumulate local evidence of symmetry through a voting procedure have been successfully used for detecting geometric symmetry in shapes. We extend such a technique to scalar fields and use it to detect geometrically symmetric regions in synthetic as well as real-world datasets. Identifying symmetry in the scalar field can significantly improve visualization and interactive exploration of the data. We demonstrate different applications of the symmetry detection method to scientific visualization: query-based exploration of scalar fields, linked selection in symmetric regions for interactive visualization, and classification of geometrically symmetric regions and its application to anomaly detection.     

Improvement of mechanical and electrical properties of epoxy resin with carbon nanofibers

In the present research, the reinforcement effect of vapor grown carbon nanofiber (VGCNF) was studied in relation to the mechanical properties and electrical conduction behavior of fabricated nanocomposites. Different weight fractions of nanofillers into epoxy resin, from 0.05 to 1 wt% and up to 2 wt% for mechanical and electrical properties were investigated. It was found that the optimum improvement in mechanical properties of nanocomposite is obtained at 0.25 wt% of carbon nanofibers. At this filler content, 23 % enhancement in tensile strength and 10 % in flexural strength have been observed. The degree of the VGCNF dispersion has been monitored by means of viscosity variation of the suspension during the sonication process to obtain the optimum sonication time. Finally, the quality of the dispersion for post-cured nanocomposites is characterized by fractured surfaces using the scanning electron microscopy. Agglomerates had a direct effect on the reduction of tensile and flexural strength of nanocomposites. The electrical conductivity was obtained by means of surface measuring method. The optimum amount of filler for the generation of a fine electrical conductivity was found to be around 0.5 wt% of VGCNF. After the threshold point, the electrical conductivity of nanocomposites was slightly raised in spite of adding more filler contents.     

An investigation on kerf characteristics in abrasive waterjet cutting of layered composites

Layered composites are “difficult-to-machine” materials as it is inhomogeneous due to the matrix properties, fibre orientation, and relative volume fraction of matrix. Abrasive waterjet cutting has proven to be a viable technique to machine such materials compared to conventional machining. This paper presents an investigation on the kerf taper angle, an important cutting performance measure, generated by abrasive waterjet (AWJ) technique to machine two types of composites: epoxy pre-impregnated graphite woven fabric and glass epoxy. Comprehensive factorial design of experiments was carried out in varying the traverse speed, abrasive flow rate, standoff distance and water pressure. Using the dimensional technique and adopting the energy conservation approach, the kerf taper angle has been related to the operating parameters in a form of a predictive model. Verification of the model for using it as a practical guideline has been found to agree with the experiments.