Electric energy forecasting domain attracts researchers due to its key role in saving resources, where mainstream existing models are based on Gradient Boosting Regression (GBR), Artificial Neural Networks (ANNs), Extreme Learning Machine (ELM) and Support Vector (SVM). These encounter high-level of non-linearity between input data output predictions limited adoptability real-world scenarios. Meanwhile, demands more robustness, higher prediction accuracy generalization ability for...

Due to industrialization and the rising demand for energy, global energy consumption has been rapidly increasing. Recent studies show that biggest portion of is consumed in residential buildings, i.e., European Union countries up 40% total by households. Most buildings industrial zones are equipped with smart sensors such as metering electric sensors, inadequately utilized better management. In this paper, we develop a hybrid convolutional neural network (CNN) an long short-term memory...

Green energy management is an economical solution for better usage, but the employed literature lacks focusing on potentials of edge intelligence in controllable Internet Things (IoT). Therefore, this article, we focus requirements todays' smart grids, homes, and industries to propose a deep-learning-based framework intelligent management. We predict future consumption short intervals time as well provide efficient way communication between distributors consumers. The key contributions...

Video anomaly recognition in smart cities is an important computer vision task that plays a vital role surveillance and public safety but challenging due to its diverse, complex, infrequent occurrence real-time environments. Various deep learning models use significant amounts of training data without generalization abilities with huge time complexity. To overcome these problems, the current work, we present efficient light-weight convolutional neural network (CNN)-based framework functional...

Scene understanding plays a crucial role in autonomous driving by utilizing sensory data for contextual information extraction and decision making. Beyond modeling advances, the enabler vehicles to become aware of their surroundings is availability visual data, which expand vehicular perception realizes awareness real-world environments. Research directions scene pursued related studies include person/vehicle detection segmentation, transition analysis, lane change, turns detection, among...

The massive amount of video data produced by surveillance networks in industries instigate various challenges exploring these videos for many applications, such as summarization (VS), analysis, indexing, and retrieval. task multiview (MVS) is very challenging due to the gigantic size data, redundancy, overlapping views, light variations, interview correlations. To address challenges, low-level features clustering-based soft computing techniques are proposed that cannot fully exploit MVS. In...

Video summarization (VS) has attracted intense attention recently due to its enormous applications in various computer vision domains, such as video retrieval, indexing, and browsing. Traditional VS researches mostly target at the effectiveness of algorithms by introducing high quality features clusters for selecting representative visual elements. Due increased density sensors network, there is a tradeoff between processing time methods with reasonable generated summaries. It challenging...

The exponential growth in the production of video contents different industries causes an urgent need for effective summarization (VS) techniques, order to get optimal storage and preservation key information video. Compared other domains, industrial videos are more challenging process, as they usually contain diverse complex events, which make their online processing a difficult task. In this article, we introduce system intelligent capturing, coarse fine redundancy removal, summary...

Nowadays, video sensors are used on a large scale for various applications, including security monitoring and smart transportation. However, the limited communication bandwidth storage constraints make it challenging to process such heterogeneous nature of Big Data in real time. Multiview summarization (MVS) enables us suppress redundant data distributed settings. The existing MVS approaches offline manner by transmitting them local or cloud server analysis, which requires extra streaming...

Fire detection and management is very important to prevent social, ecological, economic damages. However, achieving real-time fire with higher accuracy in an IoT environment a challenging task due limited storage, transmission, computation resources. To overcome these challenges, early automatic response are significant. Therefore, we develop novel framework based on lightweight convolutional neural network (CNN), requiring less training time, it applicable over resource-constrained devices....

Virtual reality (VR) has been widely used as a tool to assist people by letting them learn and simulate situations that are too dangerous risky practice in real life, one of these is road safety training for children. Traditional video- presentation-based average output results it lacks physical the involvement children during training, without any practical testing examination check learned abilities child before their exposure real-world environments. Therefore, this paper, we propose 3D...

Vision-based fire detection systems have been significantly improved by deep models; however, higher numbers of false alarms and a slow inference speed still hinder their practical applicability in real-world scenarios. For balanced trade-off between computational cost accuracy, we introduce dual attention network (DFAN) to achieve effective yet efficient detection. The first mechanism highlights the most important channels from features an existing backbone model, yielding emphasized...

Human Activity Recognition is an active research area with several Convolutional Neural Network (CNN) based features extraction and classification methods employed for surveillance other applications. However, accurate identification of HAR from a sequence frames challenging task due to cluttered background, different viewpoints, low resolution, partial occlusion. Current CNN-based techniques use large-scale computational classifiers along convolutional operators having local receptive...

Renewable energy (RE) offers major environmental and economic benefits compared to nuclear fuel-based energy; however, the data used for RE include significant randomness, intermittent behaviour, strong-volatility, hindering their integration into smart grids. Accurate prediction is a promising solution this problem can provide effective planning management services. Various predictive models have been developed improve performance better management. However, current works focus on improving...

The high-level variation of different energy generation resources makes the reliable power supply significantly challenging to end-users. These variations occur due intermittent nature output and time-varying weather conditions. recent literature focuses on improvements in consumption forecasting, which is a demand current smart grids' smooth operations with balanced amount for connected customers. Inspired by applications load therefore, this work, we develop an efficient effective hybrid...