Scenario Recognition Research Articles

Collaborative Federated Learning-Based Model for Alert Correlation and Attack Scenario Recognition

Planned and targeted attacks, such as the advanced persistent threat (APT), are highly sophisticated forms of attack. They involve numerous steps and are intended to remain within a system for an extended length of period before progressing to the next stage of action. Anticipating the next behaviors of attackers is a challenging and crucial task due to the stealthy nature of advanced attack scenarios, in addition to the possible high volumes of false positive alerts generated by different security tools such as intrusion detection systems (IDSs). Intelligent models that are capable of establishing a correlation individual between individual security alerts in order to reconstruct attack scenarios and to extract a holistic view of intrusion activities are required to exploit hidden links between different attack stages. Federated learning models performed in distributed settings have achieved successful and reliable implementations. Alerts from distributed security devices can be utilized in a collaborative manner based on several learning models to construct a federated model. Therefore, we propose an intelligent detection system that employs federated learning models to identify advanced attack scenarios such as APT. Features extracted from alerts are preprocessed and engineered to produce a model with high accuracy and fewer false positives. We conducted training on four machine learning models in a centralized learning; these models are XGBoost, Random Forest, CatBoost, and an ensemble learning model. To maintain privacy and ensure the integrity of the global model, the proposed model has been implemented using conventional neural network federated learning (CNN_FL) across several clients during the process of updating weights. The experimental findings indicate that ensemble learning achieved the highest accuracy of 88.15% in the context of centralized learning. CNN_FL has demonstrated an accuracy of 90.18% in detecting various attacks of APTs while maintaining a low false alarm rate.

Deep-Learning-Based Scenario Recognition With GNSS Measurements on Smartphones

Smartphones are in everyone’s hands for applications including navigation- and localization-based services, and scenario recognition is critical for seamless indoor and outdoor navigation. How to use smartphone sensing data to recognize different scenarios is a meaningful but challenging problem. To address this issue, we propose a structured grid-based deep-learning scenario recognition technique that uses smartphone global navigation satellite system (GNSS) measurements (satellite position, pseudorange, Doppler shift, and C/N0). In this work, the scenarios are grouped into four categories: deep indoors, shallow indoors, semioutdoors, and open outdoors. The proposed approach utilizes Voronoi tessellations to obtain structured-grid representations from satellite positions and performs computations using convolutional neural networks (CNNs) and convolutional long short-term memory (ConvLSTM) networks. With only spatial information being considered, the CNN model is used to extract the features of Voronoi tessellations for scenario recognition, achieving a high accuracy of 98.82%. Then, to enhance the robustness of the algorithm, the ConvLSTM network is adopted, which treats the measurements as spatiotemporal sequences, improving the accuracy to 99.92%. Compared with existing methods, the proposed algorithm is simple and efficient, using only GNSS measurements without the need for additional sensors. Furthermore, the latencies of the CNN and ConvLSTM models on a CPU are only 16.82 and 27.94 ms, respectively. Therefore, the proposed algorithm has the potential for real-time applications.

A Novel Algorithm for Scenario Recognition Based on MEMS Sensors of Smartphone.

The scenario is very important to smartphone-based pedestrian positioning services. The smartphone is equipped with MEMS(Micro Electro Mechanical System) sensors, which have low accuracy. Now, the methods for scenario recognition are mainly machine-learning methods. The recognition rate of a single method is not high. Multi-model fusion can improve recognition accuracy, but it needs to collect many samples, the computational cost is high, and it is heavily dependent on feature selection. Therefore, we designed the DT-BP(decision tree-Bayesian probability) scenario recognition algorithm by introducing the Bayesian state transition model based on experience design in the decision tree. The decision-tree rules and state transition probability assignment methods were respectively designed for smartphone mode and motion mode. We carried out experiments for each scenario and compared them with the methods in the references. The results showed that the method proposed in this paper has a high recognition accuracy, which is equivalent to the accuracy of multi-model machine learning, but it is simpler, easier to implement, requires less computation, and requires fewer samples.

MobileNet and Knowledge Distillation-Based Automatic Scenario Recognition Method in Vehicle-to-Vehicle Systems

Automatic scenario recognition (ASR) based on channel status information (CSI) is an important auxiliary technology for various wireless communication systems, especially vehicle-to-vehicle (V2V) systems. CSI-based ASR methods, only based on the on-board wireless receivers rather than other expensive sensors, can identify the surrounding environments of vehicles to serve assisted and autonomous driving technologies. Recently, deep learning (DL)-based ASR methods have been proposed, which generally outperform traditional, but they generally have high computation complexity and large model sizes, which are hardly introduced into mobile edge devices, such as smart vehicles. In this paper, a lightweight convolutional neural network design method, inspired by MobileNet, is introduced into ASR for reducing computation complexity and model sizes. In addition, this method may bring about performance loss, and knowledge distillation (KD) is adopted to make up for this performance loss. Simulation results demonstrated that our proposed ASR method just has 30% computation complexity and model size of the original method, and its performance loss is very limited, especially after KD.

BIOS-Based Server Intelligent Optimization.

Servers are the infrastructure of enterprise applications, and improving server performance under fixed hardware resources is an important issue. Conducting performance tuning at the application layer is common, but it is not systematic and requires prior knowledge of the running application. Some works performed tuning by dynamically adjusting the hardware prefetching configuration with a predictive model. Similarly, we design a BIOS (Basic Input/Output System)-based dynamic tuning framework for a Taishan 2280 server, including dynamic identification and static optimization. We simulate five workload scenarios (CPU-instance, etc.) with benchmark tools and perform scenario recognition dynamically with performance monitor counters (PMCs). The adjustable configurations provided by Kunpeng processing reach . Therefore, we propose a joint BIOS optimization algorithm using a deep -network. Configuration optimization is modeled as a Markov decision process starting from a feasible solution and optimizing gradually. To improve the continuous optimization capabilities, the neighborhood search method of state machine control is added. To assess its performance, we compare our algorithm with the genetic algorithm and particle swarm optimization. Our algorithm shows that it can also improve performance up to 1.10× compared to experience configuration and perform better in reducing the probability of server downtime. The dynamic tuning framework in this paper is extensible, can be trained to adapt to different scenarios, and is more suitable for servers with many adjustable configurations. Compared with the heuristic intelligent search algorithm, the proposed joint BIOS optimization algorithm can generate fewer infeasible solutions and is not easily disturbed by initialization.

QUALITY CONTROL OF LOGS IN SEMI-MECHANIZED HARVEST

Qualitative logging must be carried out on the semi-mechanized harvest to result in correct logs for the standard measures. This avoids waste and meets the demand of the clients, so management and monitoring of the execution of the logging activity are paramount. The objective of this study was to develop and test in loco a method to control the assertiveness of the lengths of logs, sectioned in a semi-mechanized manner. The method was applied in a stand of Eucalyptus sp., with a 20- year forest management cycle. The survey was divided into four action plans, with biweekly (PQ0, PQ1, PQ2, PQ3 e PQ4) monitoring of activities where PQ0 was for scenario recognition, and PQ1 to PQ4 were for the application of actions and measurement of logs and the classification according to levels (correct, acceptable, alarming, critical and unacceptable). It was observed the improvement of the percentages of the correct and acceptable levels, already in the first fortnight. The control method aided in minimizing failures and waste. It was found that after the application of the method avoided the loss of 0.73% of the total volume of the harvested and processed wood.

A dynamic scenario‐driven technique for stock price prediction and trading

AbstractIt has always been a challenge to accurately forecast the behavior of a stock market due to its extremely nonlinear and dynamic nature. Numerous studies have shown that technical indicators describing stocks in conjunction with machine learning models can serve as useful tools for forecasting in the stock market. There are various challenges, and one of them is the choice of the right technical indicators and prediction models. It is believed that there is no optimal set of technical indicators that work well in all market scenarios in a dynamic environment such as the stock market. The statement also applies to different prediction models. There is no definite winner, and different settings can emerge as winners in different market scenarios. On this premise, we propose DSdT: a dynamic scenario‐driven technique for stock price prediction and trading strategy enhancement. The proposed novel technique uses the scenario recognition and integration module to identify and integrate the current market scenario into the forecasting pipeline, resulting in a scenario‐driven stock price prediction. We use a large set of technical indicators and a shallow neural network equipped with a gating mechanism to capture and integrate the current market scenario in the prediction process. Experiments are performed on 11 stocks of the Indian Stock Market. The proposed approach yields mean absolute percentage error (MAPE) of 1.67%compared with 2.4%of its closest nonscenario‐driven counterpart for the next day's stock price prediction task. A trading strategy is also devised using the proposed technique, and the returns are compared with different baselines. Results show that the devised trading strategy yields an approximate average return of 54%compared with 25%of the return obtained by the nearest benchmark.

Detecting development scenarios of dynamic events in electric power network smart-grids.Part two“Selective Protection”

When implementingdevelopmentconcepts inthe electric power industry(such as“Smart Grid”, “Digital substation”and“Outsourcing of services”), the task of ensuring stablerelay protection operationsand automation devices is urgent. The problem is solved according to the developed structural-information (SI) method. A method for selective search of theoptimal amount of structured information for automatic decision-making is proposed. The article discusses an algorithm for recognising scenarios for the development of semantic events, which is included in the SP-method. The algorithm is applied uniformly for all hierarchical levels of recognition, based on the goals of decision making at the senior level. Control of the sequence of information events is performed in the dynamics of the passage of events along one path from all relationships according to the structural-informationmodel.Part 1 shows a collaborative structural-information model consisting of a shaping tree in a dynamic object and a recognition tree in devices. A theoretical description of the algorithm is given using the amplitude and time (Ξ,Η) selectivity windows in the general structural scheme of S-detection. The application of the method for different hierarchical levels of recognition is shown.The decision-making results are presentedin two forms, by means of a single semantic signal to indicate a group of resultsandfilling in the table of the sequence of occurrence of the recognised elementary information components. Part 2 shows the application of the SP-method at different hierarchical levels of recognition for the synthesis of a selective relay, which implements an algorithm for finding a damaged network section with single-phase ground faults in 6-35 kV distribution networks with a Petersen’scoil. The reasons for the unstable operation of algorithms of known selective relays are indicated, based on the concepts of scenario recognition. The improvement of the structure of a selective relay operating on the basis of the criterion for monitoring the coincidence of the firsthalf-waves of the mid-frequency components in the signals of transient processes is considered. Examples of the synthesis of elementary detectors of absolute,relative and cumulative actions in relation to a selective relay are given, which make it possible to fill the amount of information for general S-detection.The simulation of the operation of the synthesisedS-detector on the signals of real emergency files of the natural development of damage to the isolation of the network phase and simulation of artificial scenarios of events in the mathematical SI-modelare carried out.

Indoor Positioning Method Using WiFi RTT Based on LOS Identification and Range Calibration

WiFi-based indoor positioning methods have attracted extensive attention due to the wide installation of WiFi access points (APs). Recently, the WiFi standard was modified and introduced into a new two-way approach based on round trip time (RTT) measurement, which brings some changes for indoor positioning based on WiFi. In this work, we propose a WiFi RTT positioning method based on line of sight (LOS) identification and range calibration. Given the complexity of the indoor environment, we design a non-line of sight (NLOS) and LOS identification algorithm based on scenario recognition. The positioning scenario is recognized to assist NLOS and LOS distances identification, and gaussian process regression (GPR) is utilized to construct the scenario recognition model. Meanwhile, the calibration model for LOS distance is presented to correct the measuring distance and the scenario information is utilized to constrain the estimated position. When there is a positioning request, the positioning scenario is identified with the scenario recognition model, and LOS measuring distance is obtained based on the recognized scenario. The LOS range measurements are first calibrated and then utilized to estimate the position of the smartphone. Finally, the positioning scenario is used to constrain the estimation location to avoid it beyond the scenario. The experimental results show that the positioning effect of the proposed method is far better than that of the Least Squares (LS) algorithm, achieving a mean error (ME) of 0.862 m and root-mean-square error (RMSE) of 0.989 m.

Adversarial unsupervised domain adaptation for cross scenario waveform recognition

Deep Learning based waveform recognition has become a topical area of research recently. In fact, due to complex and heterogeneous scenarios, it is impossible to train a universal model for all waveform recognition tasks. Unsupervised domain adaptation (UDA) has an ability to take advantage of knowledge from an available source dataset and apply it to an unlabeled target dataset. Therefore, we propose a novel waveform recognition method based on an adversarial UDA, which incorporates adversarial learning to improve the cross scenario recognition performance. Specifically, a two-channel convolutional neural network is well designed for waveform recognition task and the system gradually converges an optimized stable state via two innovative methods, namely Adversarial unsupervised domain adaptation with gradient reversal (GR-AUDA) and Feature augmentation adversarial unsupervised domain adaptation (FA-AUDA). Furthermore, we sampled a more abundant waveform dataset under various scenarios, including typical indoor, outdoor, and corridor cases, using software defined radio. We evaluate our method on both public datasets and our own datasets. The results demonstrate that our proposed adversarial UDA framework is able to significantly improve the recognition performance of target domain waveforms.

Recurrent neural network based scenario recognition with Multi-constellation GNSS measurements on a smartphone

As an upper layer context-aware mobile application, fast and accurate scenario recognition is essential for seamless indoor and outdoor localization and robust positioning in complex environments. With the popularity of multi-constellation smartphones, scenario recognition based on smartphone Global Navigation Satellite System (GNSS) measurements becomes desirable. In this paper, we divide the complex environments into four categories (deep indoor, shallow indoor, semi-outdoor and open outdoor) and conduct research work in two areas. Firstly, we analyze in detail the influence of multi-constellation satellite signals on scenario recognition performance based on a Hidden Markov Model (HMM) algorithm. The experimental results show that the scenario recognition accuracy is improved significantly with the increase of the number of constellations received by smartphones. Secondly, in order to solve the description degradation of the traditional model caused by scenario transitions and environmental changes around the scenario, we propose a new scenario recognition method based on Recurrent Neural Network (RNN). Considering the computational complexity and the availability of feature values, we utilize the position-independent features as the input of the RNN model, and then evaluate the performance of the model using test sets from the new places. The results indicate that our proposed algorithm has high recognition accuracy in both isolated scenarios and transition regions, with the overall accuracy of 98.65%. Especially in the scenario transitions, the recognition accuracy reaches 90.94% and in the three times of correct recognition for scenario transitions (four times in total), the maximum transition delay is only 3 s.

Infectious Spondylitis Mimics: Mechanisms of Disease and Imaging Findings

Infectious processes of the spine are on the rise; in this scenario recognition of entities imitating infection is very important. The discovertebral unit is regarded as one of the most important and active sites in the spine. Importantly, the vertebral bony rim and the anterior corners of the vertebral bodies have significant vascularization, they are the last regions to ossify in the developmental process and suffer mechanical forces. Early septic or aseptic discitis-osteomyelitis, properly called spondylitis, involves these anterior regions. Early degeneration is characterized by disc desiccation; however, there is preferential involvement for the corners of the vertebral bodies as well. Many entities to include degenerative changes, inflammatory spondyloarthropathies, neuropathic spine, or pseudo arthrosis, among others, affect the discovertebral unit and can imitate infection. With some exceptions, important imaging findings for the identification of an infectious mimic include the absence of soft tissue enhancement or fluid collections in the paraspinal or epidural regions, and the involvement of multiple levels or the posterior elements. We review developmental, anatomical, and pathologic concepts correlating with imaging clues. Overall, our goal is to increase awareness and to improve recognition of mimicking entities.

Probabilistic Hierarchical Model Using First Person Vision for Scenario Recognition

Smart homes are becoming a growing need to prepare for a comfortable life style for the elderly and make things easy for the caretakers of the future. One important component of these systems is to identify the human activities and scenarios. As the wireless technologies are becoming advanced, they are being used to provide a low-cost, non-intrusive and privacy-conscious solution to activity recognition. However, in more complicated environments, we need to identify scenarios with subtle cues e.g. eye gaze. These situations call for a complementary vision-based solution, and we present a robust scenario recognition system by following the objects seen in eye gaze trajectories. In this paper, we present a probabilistic hierarchical model for scenario recognition using the environmental elements like objects in the scene. We utilize the fact that any scenario can be divided into constituent tasks and activities recursively to the level of atomic actions and objects. Considering bottom-up, the scenario recognition problem can be hierarchically solved by identifying the objects seen and combining them together to form coarse-grained higher level activities. This is a novel contribution to be able to recognize complete scenarios only on the basis of objects seen. We performed experiments on standard datasets of Georgia Tech Egocentric Activities (GTEA-Gaze) and unconstrained videos collected “in the Wild”; and trained an Artificial Neural Network to get a precision of 73.84% and accuracy of 92.27%.

Active Compressive Sensing via Pyroelectric Infrared Sensor for Human Situation Recognition

Conventional pyroelectric infrared (PIR) motion sensors use paired elements for the detection of moving targets. This method makes them incapable of measuring thermal signals from static targets. We need an active sensor that can detect static thermal subjects. This paper presents our design of active PIR sensors. The proposed PIR sensing systems can actively detect static thermal targets by using three methods that are suitable to different applications: 1) a sensor that can be rotated by a self-controlled servo motor for the detection of moving or static thermal subjects nearby; 2) a sensor that is equipped with a mask for low-complexity posture recognition; and 3) a sensor that can be worn on the wrist for the recognition of surrounding subjects (this sensor is especially useful for blind users). Compressive sensing (CS) theory indicates that random down-sampling method can capture more accurate information of the original signal than the evenly spaced sampling. Based on CS theory, we have developed the random sampling structures for the active PIR systems, and have built a statistical feature space for human scenario recognition. The experimental results demonstrate that the active sensing system can efficiently measure the static thermal targets, and the random sampling scheme has a better recognition performance than the even sampling scheme.

Concept generalization and fusion for abstractive sentence generation

Text summarization is either extractive or abstractive. Extractive summarization is to select the most salient pieces of information (words, phrases, and/or sentences) from a source document without adding any external information. Abstractive summarization allows an internal representation of the source document so as to produce a faithful summary of the source. In this case, external text can be inserted into the generated summary. Because of the complexity of the abstractive approach, the vast majority of work in text summarization has adopted an extractive approach.In this work, we focus on concepts fusion and generalization, i.e. where different concepts appearing in a sentence can be replaced by one concept which covers the meanings of all of them. This is one operation that can be used as part of an abstractive text summarization system. The main goal of this contribution is to enrich the research efforts on abstractive text summarization with a novel approach that allows the generalization of sentences using semantic resources. This work should be useful in intelligent systems more generally since it introduces a means to shorten sentences by producing more general (hence abstractions of the) sentences. It could be used, for instance, to display shorter texts in applications for mobile devices. It should also improve the quality of the generated text summaries by mentioning key (general) concepts. One can think of using the approach in reasoning systems where different concepts appearing in the same context are related to one another with the aim of finding a more general representation of the concepts. This could be in the context of Goal Formulation, expert systems, scenario recognition, and cognitive reasoning more generally.We present our methodology for the generalization and fusion of concepts that appear in sentences. This is achieved through (1) the detection and extraction of what we define as generalizable sentences and (2) the generation and reduction of the space of generalization versions. We introduce two approaches we have designed to select the best sentences from the space of generalization versions. Using four NLTK11The Natural Language Tool Kit (http://nltk.org/) corpora, the first approach estimates the “acceptability” of a given generalization version. The second approach is Machine Learning-based and uses contextual and specific features. The recall, precision and F1-score measures resulting from the evaluation of the concept generalization and fusion approach are presented.

Temporal and Hierarchical HMM for Activity Recognition Applied in Visual Medical Monitoring using a Multi-Camera System

We address in this paper an improved medical monitoring system through an automatic recognition of human activity in Intensive Care Units (ICUs). A multi camera vision system approach is proposed to collect video sequence for automatic analysis and interpretation of the scene. The latter is performed using Hidden Markov Model (HMM) with explicit state duration combine at the management of the hierarchical structure of the scenario. Significant experiments are carried out on the proposed monitoring system in a hospital's cardiology section in order to prove the need for computer-aided patient supervision to help clinicians in the decision making process. Temporal and hierarchical HMM handles explicitly the state duration and then provides a suitable solution for the automatic recognition of temporal events. Finally, the use of Temporal HMM (THMM) based approach improves the scenario recognition performance compared to the result of standard HMM models. Nous proposons dans cet article une solution pour améliorer le système actuel de surveillance médicale en Unité de Soins Intensifs (USIs) cardiologique grâce à un système de reconnaissance automatique d'activités humaines. Une approche de vidéo surveillance multicaméras est proposée à cet effet et permet l'acquisition des données pour l'analyse et l'interprétation automatique de la scène. Cette dernière est basée sur le Modèle de Markov Caché (MMC) avec une durée d'état explicite et intégrant une gestion de la structure hiérarchique interne des scénarios. Plusieurs séries d'expérimentations sont effectuées sur le nouveau système de surveillance proposé en USIs et démontre ainsi la nécessité d'une surveillance assistée par ordinateur des patients afin d'aider les médecins surveillants et les cliniciens dans le processus de prise de décision. De plus, le MMC temporel offre une solution très adaptée pour la reconnaissance automatique des événements en USIs. Enfin, les résultats obtenus avec le modèle de MMC temporel et hiérarchique ont été comparés à ceux des MMC classiques.

Mobile Target Scenario Recognition Via Low-Cost Pyroelectric Sensing System: Toward a Context-Enhanced Accurate Identification

Distributed binary pyroelectric sensor network (PSN) is a low-cost alternative to video systems for human monitoring applications. This paper presents a PSN-based mobile target recognition system, which aims to achieve multitarget, complex scenario recognition. In this system, a novel pseudorandom visibility mode is designed for the sensor arrays to help capture statistical information of scenarios, and a sensor array fusion scheme is adopted to facilitate discriminative feature extraction. Moreover, we propose a statistical subspace representation model called probabilistic nonnegative matrix factorization (PNMF) to seek the scenario patterns rather than the object characteristics. We also further prove that our PNMF model is a generic model for NMF based algorithms. Original NMF, sparse NMF, and smooth NMF are special cases of the PNMF model. The simulation and experimental results demonstrate the advantages of our proposed method. Our system can be further developed to function as an independent facility for intelligent monitoring applications, especially under poor illumination circumstances.

Fuzzy Logic-Based Scenario Recognition from Video Sequences

In recent years, video surveillance and monitoring have gained importance because of security and safety concerns.Banks, borders, airports, stores, and parking areas are the important application areas. There are two main parts inscenario recognition: Low level processing, including moving object detection and object tracking, and featureextraction. We have developed new features through this work which are RUD (relative upper density), RMD (relativemiddle density) and RLD (relative lower density), and we have used other features such as aspect ratio, width, height,and color of the object. High level processing, including event start-end point detection, activity detection for eachframe and scenario recognition for sequence of images. This part is the focus of our research, and different patternrecognition and classification methods are implemented and experimental results are analyzed. We looked intoseveral methods of classification which are decision tree, frequency domain classification, neural network-basedclassification, Bayes classifier, and pattern recognition methods, which are control charts, and hidden Markov models.The control chart approach, which is a decision methodology, gives more promising results than other methodologies.Overlapping between events is one of the problems, hence we applied fuzzy logic technique to solve this problem.After using this method the total accuracy increased from 95.6 to 97.2.

Temporal reasoning for scenario recognition in video-surveillance using Bayesian networks

The authors propose a high-level scenario recognition algorithm for video sequence interpretation. The recognition of scenarios is based on a Bayesian networks approach. The model of a scenario contains two main layers. The first one allows events from the observed visual features to be highlighted and the second layer is focused on the temporal reasoning stage. The temporal layer uses specific nodes permitting an event-based approach. These nodes focus on the lifetime of events highlighted from the results of the first layer. The temporal layer then estimates the qualitative and quantitative relations between the different temporal events helpful for the recognition task. The global recognition algorithm is illustrated over real indoor image sequences of an abandoned baggage scenario.

SCENARIO RECOGNITION IN MODERN BUILDING AUTOMATION

Modern building automation has to deal with very different types of demands, depending on the use of the building and therefore the persons acting within this building. To meet the demands of situation awareness in modern building automation, scenario recognition becomes more and more important to detect such demands and react to them. Two concepts of scenario recognition and their implementation will be introduced, one based on predefined templates and the other using an unsupervised learning algorithm using statistical methods. Implemented applications will be described and their advantages and disadvantages outlined. Copyright © 2007 IFAC