Problem Of Event Detection Research Articles

Avoiding Post-Processing With Event-Based Detection in Biomedical Signals.

Finding events of interest is a common task in biomedical signal processing. The detection of epileptic seizures and signal artefacts are two key examples. Epoch-based classification is the typical machine learning framework to detect such signal events because of the straightforward application of classical machine learning techniques. Usually, post-processing is required to achieve good performance and enforce temporal dependencies. Designing the right post-processing scheme to convert these classification outputs into events is a tedious, and labor-intensive element of this framework. We propose an event-based modeling framework that directly works with events as learning targets, stepping away from ad-hoc post-processing schemes to turn model outputs into events. We illustrate the practical power of this framework on simulated data and real-world data, comparing it to epoch-based modeling approaches. We show that event-based modeling (without tailored post-processing) performs on par with or better than epoch-based modeling with extensive post-processing. These results show the power of treating events as direct learning targets, instead of using ad-hoc post-processing to obtain them, severely reducing design effort. Significance The event-based modeling framework can easily be applied to other event detection problems in signal processing, removing the need for intensive task-specific post-processing.

MRME-Net: Towards multi-semantics learning and long-tail problem of efficient event detection from social messages

Discovering trending social events (e.g., major meetings, political scandals, natural disasters, etc.) from social messages is vital because it emphasizes important events and can help people comprehend the world. However, the heterogeneous semantics enrichment, severe long-tail problems, and sparse text contents of social messages pose great challenges to event detection, often leading to limited generalization ability and accuracy. In this paper, we propose a novel Multi-Relational Meta-Enhanced Network (MRME-Net) architecture to learn social events. First, we model social messages into a multi-relational message graph, incorporating abundant meta-semantics along with various meta-relations. Second, we present a multi-relational graph attention network based on Sophia by using a dual-step message aggregation mechanisms to capture the local features of neighboring messages and global semantics of mutiple relations and ultimately learn social message embeddings. We use Sophia optimizer to reduce the massive time and cost of training. Third, in order to address the long-tail problem, we introduce a locally-adapted meta-learning framework in social event detection for the first time and propose a novel META-TAILENH embedding enhancement strategy to refine tail node embeddings in multi-relational graph. Eventually, we conduct the detection of social events according to the hierarchical clustering algorithm. Extensive experiments have been carried out to evaluate MRME-Net on the MAVEN and Twitter dataset, revealing a notable improvement of 3 %–13 %, 4 %–20 % and 6 %–30 % increases on NMI, AMI and ARI in the social event detection task.

Fisher matrix based fault detection for PMUs data in power grids

In this paper, using the data collected from phasor measurement units (PMUs), two methods based on Fisher random matrix are proposed to detect faults in power grids. Firstly, the fault detection matrix is constructed and the event detection problem is reformatted as a testing problem for two-sample covariance matrices, which is related with the so-called Fisher matrix. To save computing resources, the screening step of fault interval based on the test statistic is designed to check the existence of faults. Then two point-by-point methods are proposed to determine the time of the fault in the selected interval. One method detects faults by the limiting spectral distribution of the standard Fisher matrix, which can detect the faults with higher accuracy. The other method tests the faults based on the proposed test statistic , which has a faster detection speed. Compared with existing works, the simulation results illustrate that two methods proposed in this paper cost less computational time and provide a higher degree of accuracy and sensitivity.

Discrimination between icequakes and earthquakes in southern Alaska: an exploration of waveform features using Random Forest algorithm

SUMMARY This study examines the feature space of seismic waveforms often used in machine learning applications for seismic event detection and classification problems. Our investigation centres on the southern Alaska region, where the seismic record captures diverse seismic activity, notably from the calving of marine-terminating glaciers and tectonic earthquakes along active plate boundaries. While the automated discrimination of earthquakes and glacier quakes is our nominal goal, this data set provides an outstanding opportunity to explore the general feature space of regional seismic phases. That objective has applicability beyond ice quakes and our geographic region of study. We make a noteworthy discovery that features rooted in the spectral content of seismic waveforms consistently outperform statistical and temporal features. Spectral features demonstrate robust performance, exhibiting resilience to class imbalance while being minimally impacted by factors such as epicentral distance and signal-to-noise ratio. We also conduct experiments on the transferability of the model and find that transferability primarily depends on the appearance of the waveforms. Finally, we analyse misclassified events and find examples that are identified incorrectly in the original regional catalogue.

Recursive filtering for an uncertain system with wireless switching channels and fading measurements: A microseismic event detection method

In this paper, the microseismic event detection problem is analyzed for an array of microseismic systems by developing a novel recursive filtering algorithm. The state-space model of the microseismic system with parameter uncertainty is constructed on the basis of the seismic wavelet and the ambient noise. In view of the actual situation of the wireless transmission, the switching channels and fading measurements are drawn into the microseismic system. The main purpose of the addressed problem is to devise a filter so as to precisely and automatically pick the first arrival time of the microseismic P wave and S wave. In addition, an upper bound of the filtering error covariance matrix is given and then minimized by properly designing the filter parameter. Furthermore, the mean-square exponential boundedness of the filtering error is verified. Finally, an illustrative example is provided to examine the effectiveness of the algorithm on microseismic event detection.

Evaluating machine learning architectures for sound event detection for signals with variable signal-to-noise-ratios in the Beaufort Sea.

This paper explores the challenging polyphonic sound event detection problem using machine learning architectures applied to data recorded in the Beaufort Sea during the Canada Basin Acoustic Propagation Experiment. Four candidate architectures were investigated and evaluated on nine classes of signals broadcast from moored sources that were recorded on a vertical line array of hydrophones over the course of the yearlong experiment. These signals represent a high degree of variability with respect to time-frequency characteristics, changes in signal-to-noise ratio (SNR) associated with varying signal levels as well as fluctuating ambient sound levels, and variable distributions, which resulted in class imbalances. Within this context, binary relevance, which decomposes the multi-label learning task into a number of independent binary learning tasks, was examined as an alternative to the conventional multi-label classification (MLC) approach. Binary relevance has several advantages, including flexible, lightweight model configurations that support faster model inference. In the experiments presented, binary relevance outperformed conventional MLC approach on classes with the most imbalance and lowest SNR. A deeper investigation of model performance as a function of SNR showed that binary relevance significantly improved recall within the low SNR range for all classes studied.

Transforming temporal-dynamic graphs into time-series data for solving event detection problems

Transforming temporal-dynamic graphs into time-series data for solving event detection problems

An Efficient Compressive Sensing Event-Detection Scheme for Internet of Things System Based on Sparse-Graph Codes

This work studied the event-detection problem in an Internet of Things (IoT) system, where a group of sensor nodes are placed in the region of interest to capture sparse active event sources. Using compressive sensing (CS), the event-detection problem is modeled as recovering the high-dimensional integer-valued sparse signal from incomplete linear measurements. We show that the sensing process in IoT system produces an equivalent integer CS using sparse graph codes at the sink node, for which one can devise a simple deterministic construction of a sparse measurement matrix and an efficient integer-valued signal recovery algorithm. We validated the determined measurement matrix, uniquely determined the signal coefficients, and performed an asymptotic analysis to examine the performance of the proposed approach, namely event detection with integer sum peeling (ISP), with the density evolution method. Simulation results show that the proposed ISP approach achieves a significantly higher performance compared to existing literature at various simulation scenario and match that of the theoretical results.

Detecting Extreme Traffic Events Via a Context Augmented Graph Autoencoder

Accurate and timely detection of large events on urban transportation networks enables informed mobility management. This work tackles the problem of extreme event detection on large-scale transportation networks using origin-destination mobility data, which is now widely available. Such data is highly structured in time and space, but high dimensional and sparse. Current multivariate time series anomaly detection methods cannot fully address these challenges. To exploit the structure of mobility data, we formulate the event detection problem in a novel way, as detecting anomalies in a set of time-dependent directed weighted graphs. We further propose a Context augmented Graph Autoencoder (Con-GAE) model to solve the problem, which leverages graph embedding and context embedding techniques to capture the spatial and temporal patterns. Con-GAE adopts an autoencoder framework and detects anomalies via semi-supervised learning. The performance of the method is assessed on several city-scale travel-time datasets from Uber Movement, New York taxis, and Chicago taxis and compared to state-of-the-art approaches. The proposed Con-GAE can achieve an improvement in the area under the curve score as large as 0.15 over the second best method. We also discuss real-world traffic anomalies detected by Con-GAE.

Hybrid Convolutional Networks for End-to-End Event Detection in Concurrent PPG and PCG Signals Affected by Motion Artifacts.

The accurate detection of physiologically-related events in photopletismographic (PPG) and phonocardiographic (PCG) signals, recorded by wearable sensors, is mandatory to perform the estimation of relevant cardiovascular parameters like the heart rate and the blood pressure. However, the measurement performed in uncontrolled conditions without clinical supervision leaves the detection quality particularly susceptible to noise and motion artifacts. This work proposes a new fully-automatic computational framework, based on convolutional networks, to identify and localize fiducial points in time as the foot, maximum slope and peak in PPG signal and the S1 sound in the PCG signal, both acquired by a custom chest sensor, described recently in the literature by our group. The event detection problem was reframed as a single hybrid regression-classification problem entailing a custom neural architecture to process sequentially the PPG and PCG signals. Tests were performed analysing four different acquisition conditions (rest, cycling, rest recovery and walking). Cross-validation results for the three PPG fiducial points showed identification accuracy greater than 93 % and localization error (RMSE) less than 10 ms. As expected, cycling and walking conditions provided worse results than rest and recovery, however reaching an accuracy greater than 90 % and a localization error less than 15 ms. Likewise, the identification and localization error for S1 sound were greater than 90 % and less than 25 ms. Overall, this study showcased the ability of the proposed technique to detect events with high accuracy not only for steady acquisitions but also during subject movements. We also showed that the proposed network outperformed traditional Shannon-energy-envelope method in the detection of S1 sound, reaching detection performance comparable to state of the art algorithms. Therefore, we argue that coupling chest sensors and deep learning processing techniques may disclose wearable devices to unobtrusively acquire health information, being less affected by noise and motion artifacts.

Timely Reporting of Heavy Hitters Using External Memory

Given an input stream S of size N , a ɸ-heavy hitter is an item that occurs at least ɸN times in S . The problem of finding heavy-hitters is extensively studied in the database literature. We study a real-time heavy-hitters variant in which an element must be reported shortly after we see its T = ɸ N-th occurrence (and hence it becomes a heavy hitter). We call this the Timely Event Detection ( TED ) Problem. The TED problem models the needs of many real-world monitoring systems, which demand accurate (i.e., no false negatives) and timely reporting of all events from large, high-speed streams with a low reporting threshold (high sensitivity). Like the classic heavy-hitters problem, solving the TED problem without false-positives requires large space (Ω (N) words). Thus in-RAM heavy-hitters algorithms typically sacrifice accuracy (i.e., allow false positives), sensitivity, or timeliness (i.e., use multiple passes). We show how to adapt heavy-hitters algorithms to external memory to solve the TED problem on large high-speed streams while guaranteeing accuracy, sensitivity, and timeliness. Our data structures are limited only by I/O-bandwidth (not latency) and support a tunable tradeoff between reporting delay and I/O overhead. With a small bounded reporting delay, our algorithms incur only a logarithmic I/O overhead. We implement and validate our data structures empirically using the Firehose streaming benchmark. Multi-threaded versions of our structures can scale to process 11M observations per second before becoming CPU bound. In comparison, a naive adaptation of the standard heavy-hitters algorithm to external memory would be limited by the storage device’s random I/O throughput, i.e., ≈100K observations per second.

Smart Director: An Event-Driven Directing System for Live Broadcasting

Live video broadcasting normally requires a multitude of skills and expertise with domain knowledge to enable multi-camera productions. As the number of cameras keeps increasing, directing a live sports broadcast has now become more complicated and challenging than ever before. The broadcast directors need to be much more concentrated, responsive, and knowledgeable, during the production. To relieve the directors from their intensive efforts, we develop an innovative automated sports broadcast directing system, called Smart Director, which aims at mimicking the typical human-in-the-loop broadcasting process to automatically create near-professional broadcasting programs in real-time by using a set of advanced multi-view video analysis algorithms. Inspired by the so-called “three-event” construction of sports broadcast [ 14 ], we build our system with an event-driven pipeline consisting of three consecutive novel components: (1) the Multi-View Event Localization to detect events by modeling multi-view correlations, (2) the Multi-View Highlight Detection to rank camera views by the visual importance for view selection, and (3) the Auto-Broadcasting Scheduler to control the production of broadcasting videos. To our best knowledge, our system is the first end-to-end automated directing system for multi-camera sports broadcasting, completely driven by the semantic understanding of sports events. It is also the first system to solve the novel problem of multi-view joint event detection by cross-view relation modeling. We conduct both objective and subjective evaluations on a real-world multi-camera soccer dataset, which demonstrate the quality of our auto-generated videos is comparable to that of the human-directed videos. Thanks to its faster response, our system is able to capture more fast-passing and short-duration events which are usually missed by human directors.

Fault Detection Method and Simulation Based on Abnormal Data Analysis in Wireless Sensor Networks

With the rapid development of Internet of things and information technology, wireless sensor network technology is widely used in industrial monitoring. However, limited by the architecture characteristics, software and hardware characteristics, and complex external environmental factors of wireless sensor networks, there are often serious abnormalities in the monitoring data of wireless sensor networks, which further affect the judgment and response of users. Based on this, this paper optimizes and improves the fault detection algorithm of related abnormal data analysis in wireless sensor networks from two angles and verifies the algorithm at the same time. In the first level, aiming at the problem of insufficient spatial cooperation faced by the network abnormal data detection level, this paper first establishes a stable neighbor screening model based on the wireless network and filters and analyzes the reliability of the network cooperative data nodes and then establishes the detection data stability evaluation model by using the spatiotemporal correlation corresponding to the data nodes. Realize abnormal data detection. On the second level, aiming at the problem of wireless network abnormal event detection, this paper proposes a spatial clustering optimization algorithm, which mainly clusters the detection data flow in the wireless network time window through the clustering algorithm, and analyzes the clustering data, so as to realize the detection of network abnormal events, so as to retain the characteristics of events and further classify the abnormal data events. This paper will verify the realizability and superiority of the improved optimization algorithm through simulation technology. Experiments show that the fault detection rate based on abnormal data analysis is as high as 97%, which is 5% higher than the traditional fault detection rate. At the same time, the corresponding fault false detection rate is low and controlled below 1%. The efficiency of this algorithm is about 10% higher than that of the traditional algorithm.

Spotting Football Events Using Two-Stream Convolutional Neural Network and Dilated Recurrent Neural Network

This paper addresses the problem of event detection and localization in long football (soccer) videos. Our key idea is that understanding the long-range dependencies between video frames is imperative for accurate event localization in long football videos. Additionally, proper event detection is not likely for fast movements in football videos without considering mid-range and short-range correlations between neighboring video frames. We argue that event spotting can be considerably improved by considering short-range to long-range frame dependencies in a unified architecture. To model long-range and mid-range dependencies, we propose to use the dilated recurrent neural network (DilatedRNN) with long short-term memory (LSTM) units, grounded on two-stream convolutional neural network (Two-stream CNN) features. While two-stream CNN extracts local spatiotemporal features necessary for fine-level details, the DilatedRNN makes the information obtained from distant frames available for the classifier and spotting algorithms. Evaluating our event spotting algorithm on the largest publicly available benchmark football dataset –SoccerNet– shows an accuracy improvement of 0.8% - 13.6% compared to state of the art, and up to 30.1% accuracy gain in comparison to the baselines. We also investigate the contribution of each neural network component in spotting accuracy through an extensive ablation study.

Hierarchical Regulated Iterative Network for Joint Task of Music Detection and Music Relative Loudness Estimation

One practical requirement of the music copyright management is the estimation of music relative loudness, which is mostly ignored in existing music detection works. To solve this problem, we study the joint task of music detection and music relative loudness estimation. To be specific, we observe that the joint task has two characteristics, i.e., temporality and hierarchy, which could facilitate to obtain the solution. For example, a tiny fragment of audio is temporally related to its neighbor fragments because they may all belong to the same event, and the event classes of the fragment in the two tasks have a hierarchical relationship. Based on the above observation, we reformulate the joint task as hierarchical event detection and localization problem. To solve this problem, we further propose Hierarchical Regulated Iterative Networks (HRIN), which includes two variants, termed as HRIN-r and HRIN-cr, which are based on recurrent and convolutional recurrent modules. To enjoy the joint task's characteristics, our models employ an iterative framework to achieve encouraging capability in temporal modeling while designing three hierarchical violation penalties to regulate hierarchy. Extensive experiments on the currently largest dataset (i.e., OpenBMAT) show that the promising performance of our HRIN in the segment-level and event-level evaluations.

An empirical exploration of performance metrics for event detection algorithms in Non-Intrusive Load Monitoring

The field of Non-Intrusive Load Monitoring (NILM) gained prominence due to its promise of inferring the energy consumption of individual appliances by analyzing only the aggregated consumption. Still, despite some research efforts towards producing meaningful comparisons between approaches, it is not yet possible to find a proven and formally accepted set of metrics to do this. Against this background, this paper focuses on understanding the challenges of defining a consistent set of performance metrics for this problem. More concretely, it reports on an empirical exploration of 23 performance metrics’ behavior when applied to event-detection algorithms, identifying relationships and clusters between them. The results indicate that when applied to this problem, the performance metrics will show some considerable differences in behavior compared to other, more traditional, machine-learning domains. The results also suggest that most of the differences occur due to the unbalanced nature of the event detection problem, in which the number of positive cases (True Positives and True Negatives) is much higher than the number of false situations (False Positives and False Negatives). Furthermore, the results suggest that additional research is needed to find proper domain-specific performance metrics that take full consideration of the properties of the aggregated load.

Leveraging burst in twitter network communities for event detection

Detecting protest events using social media is an important task with many useful applications to emergency services, law enforcement agencies, and other stakeholders. A plethora of research on event detection using social media has presented myriad approaches relying on tweet contents (text) to solve the event detection problem, with notable improvements over time. Despite the myriad of existing research, the use of the structural relationships among users in online Twitter network communities for event detection is rarely observed. In this work, we present a novel protest event detection framework called SensorTree. SensorTree utilizes the network structural connections among users in a community for protest event detection. The SensorTree framework tracks information propagation in Twitter network communities to model the sudden change in growth of these communities as burst for event detection. Once burst is detected, SensorTree builds a tensorized topic model to extract events. To show the prowess of SensorTree for event detection, we conduct extensive experiments on geographically diverse Twitter datasets using qualitative and quantitative evaluations. We further show the superiority of SensorTree by comparing our results to several existing state-of-the-art methods. SensorTree outperforms the baselines as well as the comparison models. The results further suggest that utilizing network community structure yields concise and accurate event detection. We also present case studies on real-world protest event to further show that SensorTree is capable of detecting events with fine granularity description without any language restrictions.

Leveraging Phase Transition of Topics for Event Detection in Social Media

With the advancement of technology, many processes in our world have been reformulated, updated, and digitized. Therefore, interpersonal relationships have also been following this trend so that social networks have become increasingly present in our lives. Given this context, social network users create and share a large amount of data, from content about their daily lives, funny facts, as well as information about traffic, weather, and various subjects. The problem of event detection in social media, such as Twitter, is related to the identification of the first story on a topic of interest. In this work, we propose a novel approach based on the observation that tweets are subjected to a continuous phase transition when an event takes place, i.e., its underlying dynamic changes. Our proposal consists of a formal characterization of the phase transition that occurs when an event takes place, and the use of this characterization to devise a new method to detect events in Twitter, based on calculating the entropy of the keywords extracted from the content of tweets (regardless of the language used). We evaluated the performance of our approach using seven data sets, and we outperformed nine different techniques present in the literature. Unlike the work found in the literature, we present a theoretical rationale about the existence of phase transitions. For this, we characterize a model, already existing in the literature, of phase transitions described by differential equations, where we find correspondence between the model used in the study and the real data. The experimental results show that our proposal significantly improves the learning performance for the metrics used.

A hybrid swarm intelligence based approach for abnormal event detection in crowded environments

In this paper, we propose a hybrid swarm intelligence based approach to tackle the problem of abnormal event detection in crowded environments in surveillance videos. In the proposed approach, a video frame is subjected to a series of operations to extract most salient information from it. Initially, a novel discriminative 2D variance plane corresponding to (and equal in dimensions to) each video frame is constructed in which the value at each pixel location represents the variance of optical flow field magnitude in the local spatio-temporal neighborhood of that pixel. Consequently, a modified ant colony optimization (ACO) clustering algorithm is employed to partition the 2D variance plane into salient and non-salient clusters. The cluster with salient pixels represents the regions of a video frame where optical flow variations inside the local spatio-temporal neighborhood of a pixel are high and is selected for further computation. Finally, a novel predator-prey algorithm is implemented and predators are advected over the prey values in the selected cluster to compute histogram of swarms (HOS) for a particular frame. The proposed approach outperforms state of the art on two commonly used datasets in our experiments, i.e., UMN crowd anomaly dataset and UCF web dataset.

Lyapunov–based Anomaly Detection in Preferential Attachment Networks

Abstract Network models aim to explain patterns of empirical relationships based on mechanisms that operate under various principles for establishing and removing links. The principle of preferential attachment forms a basis for the well-known Barabási–Albert model, which describes a stochastic preferential attachment process where newly added nodes tend to connect to the more highly connected ones. Previous work has shown that a wide class of such models are able to recreate power law degree distributions. This paper characterizes the cumulative degree distribution of the Barabási–Albert model as an invariant set and shows that this set is not only a global attractor, but it is also stable in the sense of Lyapunov. Stability in this context means that, for all initial configurations, the cumulative degree distributions of subsequent networks remain, for all time, close to the limit distribution. We use the stability properties of the distribution to design a semi-supervised technique for the problem of anomalous event detection on networks.